From 4951704b4e23d71b99ac933d8e6993bc6225ac13 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 12 May 2008 03:29:11 -0700 Subject: syncppp: Fix crashes. The syncppp layer wants a mid-level netdev private pointer. It was using netdev->priv but that only worked by accident, and thus this scheme was broken when the device private allocation strategy changed. Add a proper mid-layer private pointer for uses like this, update syncppp and all users, and remove the HDLC_PPP broken tag from drivers/net/wan/Kconfig Signed-off-by: David S. Miller --- include/linux/netdevice.h | 3 +++ include/net/syncppp.h | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 7c1d4466583b..746901774d49 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -715,6 +715,9 @@ struct net_device struct net *nd_net; #endif + /* mid-layer private */ + void *ml_priv; + /* bridge stuff */ struct net_bridge_port *br_port; /* macvlan */ diff --git a/include/net/syncppp.h b/include/net/syncppp.h index 877efa434700..e43f4070d892 100644 --- a/include/net/syncppp.h +++ b/include/net/syncppp.h @@ -59,7 +59,7 @@ struct ppp_device static inline struct sppp *sppp_of(struct net_device *dev) { - struct ppp_device **ppp = dev->priv; + struct ppp_device **ppp = dev->ml_priv; BUG_ON((*ppp)->dev != dev); return &(*ppp)->sppp; } -- cgit From 94d149c34cda933ff5096aca94bb23bf68602f4e Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 12 May 2008 16:33:33 -0700 Subject: sparc: Fix mremap address range validation. Just like mmap, we need to validate address ranges regardless of MAP_FIXED. sparc{,64}_mmap_check()'s flag argument is unused, remove. Based upon a report and preliminary patch by Jan Lieskovsky Signed-off-by: David S. Miller --- arch/sparc/kernel/sys_sparc.c | 48 ++++----------------------------------- arch/sparc64/kernel/sys_sparc.c | 36 ++++------------------------- arch/sparc64/kernel/sys_sparc32.c | 33 ++------------------------- include/asm-sparc/mman.h | 5 ++-- include/asm-sparc64/mman.h | 5 ++-- 5 files changed, 15 insertions(+), 112 deletions(-) (limited to 'include') diff --git a/arch/sparc/kernel/sys_sparc.c b/arch/sparc/kernel/sys_sparc.c index e995491c4436..3c6b49a53ae8 100644 --- a/arch/sparc/kernel/sys_sparc.c +++ b/arch/sparc/kernel/sys_sparc.c @@ -219,7 +219,7 @@ out: return err; } -int sparc_mmap_check(unsigned long addr, unsigned long len, unsigned long flags) +int sparc_mmap_check(unsigned long addr, unsigned long len) { if (ARCH_SUN4C_SUN4 && (len > 0x20000000 || @@ -295,52 +295,14 @@ asmlinkage unsigned long sparc_mremap(unsigned long addr, unsigned long old_len, unsigned long new_len, unsigned long flags, unsigned long new_addr) { - struct vm_area_struct *vma; unsigned long ret = -EINVAL; - if (ARCH_SUN4C_SUN4) { - if (old_len > 0x20000000 || new_len > 0x20000000) - goto out; - if (addr < 0xe0000000 && addr + old_len > 0x20000000) - goto out; - } - if (old_len > TASK_SIZE - PAGE_SIZE || - new_len > TASK_SIZE - PAGE_SIZE) + + if (unlikely(sparc_mmap_check(addr, old_len))) + goto out; + if (unlikely(sparc_mmap_check(new_addr, new_len))) goto out; down_write(¤t->mm->mmap_sem); - if (flags & MREMAP_FIXED) { - if (ARCH_SUN4C_SUN4 && - new_addr < 0xe0000000 && - new_addr + new_len > 0x20000000) - goto out_sem; - if (new_addr + new_len > TASK_SIZE - PAGE_SIZE) - goto out_sem; - } else if ((ARCH_SUN4C_SUN4 && addr < 0xe0000000 && - addr + new_len > 0x20000000) || - addr + new_len > TASK_SIZE - PAGE_SIZE) { - unsigned long map_flags = 0; - struct file *file = NULL; - - ret = -ENOMEM; - if (!(flags & MREMAP_MAYMOVE)) - goto out_sem; - - vma = find_vma(current->mm, addr); - if (vma) { - if (vma->vm_flags & VM_SHARED) - map_flags |= MAP_SHARED; - file = vma->vm_file; - } - - new_addr = get_unmapped_area(file, addr, new_len, - vma ? vma->vm_pgoff : 0, - map_flags); - ret = new_addr; - if (new_addr & ~PAGE_MASK) - goto out_sem; - flags |= MREMAP_FIXED; - } ret = do_mremap(addr, old_len, new_len, flags, new_addr); -out_sem: up_write(¤t->mm->mmap_sem); out: return ret; diff --git a/arch/sparc64/kernel/sys_sparc.c b/arch/sparc64/kernel/sys_sparc.c index 0dbc941f130e..ac1bff58c1ac 100644 --- a/arch/sparc64/kernel/sys_sparc.c +++ b/arch/sparc64/kernel/sys_sparc.c @@ -542,8 +542,7 @@ asmlinkage long sparc64_personality(unsigned long personality) return ret; } -int sparc64_mmap_check(unsigned long addr, unsigned long len, - unsigned long flags) +int sparc64_mmap_check(unsigned long addr, unsigned long len) { if (test_thread_flag(TIF_32BIT)) { if (len >= STACK_TOP32) @@ -609,46 +608,19 @@ asmlinkage unsigned long sys64_mremap(unsigned long addr, unsigned long old_len, unsigned long new_len, unsigned long flags, unsigned long new_addr) { - struct vm_area_struct *vma; unsigned long ret = -EINVAL; if (test_thread_flag(TIF_32BIT)) goto out; if (unlikely(new_len >= VA_EXCLUDE_START)) goto out; - if (unlikely(invalid_64bit_range(addr, old_len))) + if (unlikely(sparc64_mmap_check(addr, old_len))) + goto out; + if (unlikely(sparc64_mmap_check(new_addr, new_len))) goto out; down_write(¤t->mm->mmap_sem); - if (flags & MREMAP_FIXED) { - if (invalid_64bit_range(new_addr, new_len)) - goto out_sem; - } else if (invalid_64bit_range(addr, new_len)) { - unsigned long map_flags = 0; - struct file *file = NULL; - - ret = -ENOMEM; - if (!(flags & MREMAP_MAYMOVE)) - goto out_sem; - - vma = find_vma(current->mm, addr); - if (vma) { - if (vma->vm_flags & VM_SHARED) - map_flags |= MAP_SHARED; - file = vma->vm_file; - } - - /* MREMAP_FIXED checked above. */ - new_addr = get_unmapped_area(file, addr, new_len, - vma ? vma->vm_pgoff : 0, - map_flags); - ret = new_addr; - if (new_addr & ~PAGE_MASK) - goto out_sem; - flags |= MREMAP_FIXED; - } ret = do_mremap(addr, old_len, new_len, flags, new_addr); -out_sem: up_write(¤t->mm->mmap_sem); out: return ret; diff --git a/arch/sparc64/kernel/sys_sparc32.c b/arch/sparc64/kernel/sys_sparc32.c index 1aa4288125f2..ba5bd626b39e 100644 --- a/arch/sparc64/kernel/sys_sparc32.c +++ b/arch/sparc64/kernel/sys_sparc32.c @@ -867,44 +867,15 @@ asmlinkage unsigned long sys32_mremap(unsigned long addr, unsigned long old_len, unsigned long new_len, unsigned long flags, u32 __new_addr) { - struct vm_area_struct *vma; unsigned long ret = -EINVAL; unsigned long new_addr = __new_addr; - if (old_len > STACK_TOP32 || new_len > STACK_TOP32) + if (unlikely(sparc64_mmap_check(addr, old_len))) goto out; - if (addr > STACK_TOP32 - old_len) + if (unlikely(sparc64_mmap_check(new_addr, new_len))) goto out; down_write(¤t->mm->mmap_sem); - if (flags & MREMAP_FIXED) { - if (new_addr > STACK_TOP32 - new_len) - goto out_sem; - } else if (addr > STACK_TOP32 - new_len) { - unsigned long map_flags = 0; - struct file *file = NULL; - - ret = -ENOMEM; - if (!(flags & MREMAP_MAYMOVE)) - goto out_sem; - - vma = find_vma(current->mm, addr); - if (vma) { - if (vma->vm_flags & VM_SHARED) - map_flags |= MAP_SHARED; - file = vma->vm_file; - } - - /* MREMAP_FIXED checked above. */ - new_addr = get_unmapped_area(file, addr, new_len, - vma ? vma->vm_pgoff : 0, - map_flags); - ret = new_addr; - if (new_addr & ~PAGE_MASK) - goto out_sem; - flags |= MREMAP_FIXED; - } ret = do_mremap(addr, old_len, new_len, flags, new_addr); -out_sem: up_write(¤t->mm->mmap_sem); out: return ret; diff --git a/include/asm-sparc/mman.h b/include/asm-sparc/mman.h index e18be984c01d..3d16b40bb8ef 100644 --- a/include/asm-sparc/mman.h +++ b/include/asm-sparc/mman.h @@ -24,9 +24,8 @@ #ifdef __KERNEL__ #ifndef __ASSEMBLY__ -#define arch_mmap_check sparc_mmap_check -int sparc_mmap_check(unsigned long addr, unsigned long len, - unsigned long flags); +#define arch_mmap_check(addr,len,flags) sparc_mmap_check(addr,len) +int sparc_mmap_check(unsigned long addr, unsigned long len); #endif #endif diff --git a/include/asm-sparc64/mman.h b/include/asm-sparc64/mman.h index e584563b56eb..625be4d61baf 100644 --- a/include/asm-sparc64/mman.h +++ b/include/asm-sparc64/mman.h @@ -24,9 +24,8 @@ #ifdef __KERNEL__ #ifndef __ASSEMBLY__ -#define arch_mmap_check sparc64_mmap_check -int sparc64_mmap_check(unsigned long addr, unsigned long len, - unsigned long flags); +#define arch_mmap_check(addr,len,flags) sparc64_mmap_check(addr,len) +int sparc64_mmap_check(unsigned long addr, unsigned long len); #endif #endif -- cgit From 8388e3da34edb141362bb42811ee487dfec15525 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 12 May 2008 20:17:33 -0700 Subject: net: Set LL_MAX_HEADER properly for wireless. Wireless networking, particularly with MESH enabled, has quite strong requirements for link-layer header space. Based upon some numbers and descriptions from Johannes Berg we use 96 (same as AX25) for plain wireless, and with mesh enabled we use 128. In the process, simplify the cpp conditional logic here by ordering the cases by those needing the most space down to those needing the least case. Signed-off-by: David S. Miller --- include/linux/netdevice.h | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 746901774d49..a3fb57fde623 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -93,14 +93,16 @@ struct wireless_dev; * used. */ -#if !defined(CONFIG_AX25) && !defined(CONFIG_AX25_MODULE) && !defined(CONFIG_TR) -#define LL_MAX_HEADER 32 +#if defined(CONFIG_WLAN_80211) || defined(CONFIG_AX25) || defined(CONFIG_AX25_MODULE) +# if defined(CONFIG_MAC80211_MESH) +# define LL_MAX_HEADER 128 +# else +# define LL_MAX_HEADER 96 +# endif +#elif defined(CONFIG_TR) +# define LL_MAX_HEADER 48 #else -#if defined(CONFIG_AX25) || defined(CONFIG_AX25_MODULE) -#define LL_MAX_HEADER 96 -#else -#define LL_MAX_HEADER 48 -#endif +# define LL_MAX_HEADER 32 #endif #if !defined(CONFIG_NET_IPIP) && !defined(CONFIG_NET_IPIP_MODULE) && \ -- cgit From f5184d267c1aedb9b7a8cc44e08ff6b8d382c3b5 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 12 May 2008 20:48:31 -0700 Subject: net: Allow netdevices to specify needed head/tailroom This patch adds needed_headroom/needed_tailroom members to struct net_device and updates many places that allocate sbks to use them. Not all of them can be converted though, and I'm sure I missed some (I mostly grepped for LL_RESERVED_SPACE) Signed-off-by: Johannes Berg Signed-off-by: David S. Miller --- include/linux/netdevice.h | 16 ++++++++++++++-- net/core/netpoll.c | 2 +- net/econet/af_econet.c | 2 +- net/ipv4/arp.c | 2 +- net/ipv4/igmp.c | 4 ++-- net/ipv4/ipconfig.c | 6 +++--- net/ipv4/raw.c | 10 ++++------ net/ipv6/ip6_output.c | 2 +- net/ipv6/mcast.c | 4 ++-- net/ipv6/ndisc.c | 4 ++-- net/ipv6/raw.c | 10 ++++------ net/packet/af_packet.c | 2 +- net/xfrm/xfrm_output.c | 6 +++--- 13 files changed, 39 insertions(+), 31 deletions(-) (limited to 'include') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a3fb57fde623..b11e6e19e96c 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -246,11 +246,16 @@ struct hh_cache * * We could use other alignment values, but we must maintain the * relationship HH alignment <= LL alignment. + * + * LL_ALLOCATED_SPACE also takes into account the tailroom the device + * may need. */ #define LL_RESERVED_SPACE(dev) \ - (((dev)->hard_header_len&~(HH_DATA_MOD - 1)) + HH_DATA_MOD) + ((((dev)->hard_header_len+(dev)->needed_headroom)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD) #define LL_RESERVED_SPACE_EXTRA(dev,extra) \ - ((((dev)->hard_header_len+extra)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD) + ((((dev)->hard_header_len+(dev)->needed_headroom+(extra))&~(HH_DATA_MOD - 1)) + HH_DATA_MOD) +#define LL_ALLOCATED_SPACE(dev) \ + ((((dev)->hard_header_len+(dev)->needed_headroom+(dev)->needed_tailroom)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD) struct header_ops { int (*create) (struct sk_buff *skb, struct net_device *dev, @@ -569,6 +574,13 @@ struct net_device unsigned short type; /* interface hardware type */ unsigned short hard_header_len; /* hardware hdr length */ + /* extra head- and tailroom the hardware may need, but not in all cases + * can this be guaranteed, especially tailroom. Some cases also use + * LL_MAX_HEADER instead to allocate the skb. + */ + unsigned short needed_headroom; + unsigned short needed_tailroom; + struct net_device *master; /* Pointer to master device of a group, * which this device is member of. */ diff --git a/net/core/netpoll.c b/net/core/netpoll.c index b04d643fc3c7..8fb134da0346 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -419,7 +419,7 @@ static void arp_reply(struct sk_buff *skb) return; size = arp_hdr_len(skb->dev); - send_skb = find_skb(np, size + LL_RESERVED_SPACE(np->dev), + send_skb = find_skb(np, size + LL_ALLOCATED_SPACE(np->dev), LL_RESERVED_SPACE(np->dev)); if (!send_skb) diff --git a/net/econet/af_econet.c b/net/econet/af_econet.c index 68d154480043..7c9bb13b1539 100644 --- a/net/econet/af_econet.c +++ b/net/econet/af_econet.c @@ -340,7 +340,7 @@ static int econet_sendmsg(struct kiocb *iocb, struct socket *sock, dev_hold(dev); - skb = sock_alloc_send_skb(sk, len+LL_RESERVED_SPACE(dev), + skb = sock_alloc_send_skb(sk, len+LL_ALLOCATED_SPACE(dev), msg->msg_flags & MSG_DONTWAIT, &err); if (skb==NULL) goto out_unlock; diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index 68b72a7a1806..418862f1bf22 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -570,7 +570,7 @@ struct sk_buff *arp_create(int type, int ptype, __be32 dest_ip, * Allocate a buffer */ - skb = alloc_skb(arp_hdr_len(dev) + LL_RESERVED_SPACE(dev), GFP_ATOMIC); + skb = alloc_skb(arp_hdr_len(dev) + LL_ALLOCATED_SPACE(dev), GFP_ATOMIC); if (skb == NULL) return NULL; diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 6250f4239b61..2769dc4a4c84 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -292,7 +292,7 @@ static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size) struct iphdr *pip; struct igmpv3_report *pig; - skb = alloc_skb(size + LL_RESERVED_SPACE(dev), GFP_ATOMIC); + skb = alloc_skb(size + LL_ALLOCATED_SPACE(dev), GFP_ATOMIC); if (skb == NULL) return NULL; @@ -653,7 +653,7 @@ static int igmp_send_report(struct in_device *in_dev, struct ip_mc_list *pmc, return -1; } - skb=alloc_skb(IGMP_SIZE+LL_RESERVED_SPACE(dev), GFP_ATOMIC); + skb=alloc_skb(IGMP_SIZE+LL_ALLOCATED_SPACE(dev), GFP_ATOMIC); if (skb == NULL) { ip_rt_put(rt); return -1; diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index 89dee4346f60..ed45037ce9be 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -710,14 +710,14 @@ static void __init ic_bootp_send_if(struct ic_device *d, unsigned long jiffies_d struct net_device *dev = d->dev; struct sk_buff *skb; struct bootp_pkt *b; - int hh_len = LL_RESERVED_SPACE(dev); struct iphdr *h; /* Allocate packet */ - skb = alloc_skb(sizeof(struct bootp_pkt) + hh_len + 15, GFP_KERNEL); + skb = alloc_skb(sizeof(struct bootp_pkt) + LL_ALLOCATED_SPACE(dev) + 15, + GFP_KERNEL); if (!skb) return; - skb_reserve(skb, hh_len); + skb_reserve(skb, LL_RESERVED_SPACE(dev)); b = (struct bootp_pkt *) skb_put(skb, sizeof(struct bootp_pkt)); memset(b, 0, sizeof(struct bootp_pkt)); diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 11d7f753a820..fead049daf43 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -322,7 +322,6 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, unsigned int flags) { struct inet_sock *inet = inet_sk(sk); - int hh_len; struct iphdr *iph; struct sk_buff *skb; unsigned int iphlen; @@ -336,13 +335,12 @@ static int raw_send_hdrinc(struct sock *sk, void *from, size_t length, if (flags&MSG_PROBE) goto out; - hh_len = LL_RESERVED_SPACE(rt->u.dst.dev); - - skb = sock_alloc_send_skb(sk, length+hh_len+15, - flags&MSG_DONTWAIT, &err); + skb = sock_alloc_send_skb(sk, + length + LL_ALLOCATED_SPACE(rt->u.dst.dev) + 15, + flags & MSG_DONTWAIT, &err); if (skb == NULL) goto error; - skb_reserve(skb, hh_len); + skb_reserve(skb, LL_RESERVED_SPACE(rt->u.dst.dev)); skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 0af2e055f883..48cdce9c696c 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -780,7 +780,7 @@ slow_path: * Allocate buffer. */ - if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_RESERVED_SPACE(rt->u.dst.dev), GFP_ATOMIC)) == NULL) { + if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_ALLOCATED_SPACE(rt->u.dst.dev), GFP_ATOMIC)) == NULL) { NETDEBUG(KERN_INFO "IPv6: frag: no memory for new fragment!\n"); IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS); diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c index 54f91efdae58..fd632dd7f98d 100644 --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1411,7 +1411,7 @@ static struct sk_buff *mld_newpack(struct net_device *dev, int size) IPV6_TLV_PADN, 0 }; /* we assume size > sizeof(ra) here */ - skb = sock_alloc_send_skb(sk, size + LL_RESERVED_SPACE(dev), 1, &err); + skb = sock_alloc_send_skb(sk, size + LL_ALLOCATED_SPACE(dev), 1, &err); if (!skb) return NULL; @@ -1790,7 +1790,7 @@ static void igmp6_send(struct in6_addr *addr, struct net_device *dev, int type) payload_len = len + sizeof(ra); full_len = sizeof(struct ipv6hdr) + payload_len; - skb = sock_alloc_send_skb(sk, LL_RESERVED_SPACE(dev) + full_len, 1, &err); + skb = sock_alloc_send_skb(sk, LL_ALLOCATED_SPACE(dev) + full_len, 1, &err); if (skb == NULL) { rcu_read_lock(); diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index 2c74885f8355..a55fc05b8125 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -479,7 +479,7 @@ static void __ndisc_send(struct net_device *dev, skb = sock_alloc_send_skb(sk, (MAX_HEADER + sizeof(struct ipv6hdr) + - len + LL_RESERVED_SPACE(dev)), + len + LL_ALLOCATED_SPACE(dev)), 1, &err); if (!skb) { ND_PRINTK0(KERN_ERR @@ -1521,7 +1521,7 @@ void ndisc_send_redirect(struct sk_buff *skb, struct neighbour *neigh, buff = sock_alloc_send_skb(sk, (MAX_HEADER + sizeof(struct ipv6hdr) + - len + LL_RESERVED_SPACE(dev)), + len + LL_ALLOCATED_SPACE(dev)), 1, &err); if (buff == NULL) { ND_PRINTK0(KERN_ERR diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 396f0ea11090..232e0dc45bf5 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -609,7 +609,6 @@ static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, struct ipv6_pinfo *np = inet6_sk(sk); struct ipv6hdr *iph; struct sk_buff *skb; - unsigned int hh_len; int err; if (length > rt->u.dst.dev->mtu) { @@ -619,13 +618,12 @@ static int rawv6_send_hdrinc(struct sock *sk, void *from, int length, if (flags&MSG_PROBE) goto out; - hh_len = LL_RESERVED_SPACE(rt->u.dst.dev); - - skb = sock_alloc_send_skb(sk, length+hh_len+15, - flags&MSG_DONTWAIT, &err); + skb = sock_alloc_send_skb(sk, + length + LL_ALLOCATED_SPACE(rt->u.dst.dev) + 15, + flags & MSG_DONTWAIT, &err); if (skb == NULL) goto error; - skb_reserve(skb, hh_len); + skb_reserve(skb, LL_RESERVED_SPACE(rt->u.dst.dev)); skb->priority = sk->sk_priority; skb->mark = sk->sk_mark; diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 25070240d4ae..2cee87da4441 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -743,7 +743,7 @@ static int packet_sendmsg(struct kiocb *iocb, struct socket *sock, if (len > dev->mtu+reserve) goto out_unlock; - skb = sock_alloc_send_skb(sk, len + LL_RESERVED_SPACE(dev), + skb = sock_alloc_send_skb(sk, len + LL_ALLOCATED_SPACE(dev), msg->msg_flags & MSG_DONTWAIT, &err); if (skb==NULL) goto out_unlock; diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c index 09cd9c0c2d80..3f964db908a7 100644 --- a/net/xfrm/xfrm_output.c +++ b/net/xfrm/xfrm_output.c @@ -25,11 +25,11 @@ static int xfrm_state_check_space(struct xfrm_state *x, struct sk_buff *skb) struct dst_entry *dst = skb->dst; int nhead = dst->header_len + LL_RESERVED_SPACE(dst->dev) - skb_headroom(skb); + int ntail = dst->dev->needed_tailroom - skb_tailroom(skb); - if (nhead > 0) - return pskb_expand_head(skb, nhead, 0, GFP_ATOMIC); + if (nhead > 0 || ntail > 0) + return pskb_expand_head(skb, nhead, ntail, GFP_ATOMIC); - /* Check tail too... */ return 0; } -- cgit From 9a28dbf8af11d127bf1c644143e7882cb91515dd Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 12 May 2008 22:45:15 -0700 Subject: sparc64: Use a TS_RESTORE_SIGMASK This mirrors x86 changeset 5a8da0ea82db6fa9737041381079fd16f25dcce2 ("signals: x86 TS_RESTORE_SIGMASK") on sparc64. Signed-off-by: David S. Miller --- arch/sparc64/kernel/rtrap.S | 6 +++--- arch/sparc64/kernel/signal.c | 21 +++++++++++---------- arch/sparc64/kernel/signal32.c | 13 ++++++------- include/asm-sparc64/thread_info.h | 28 ++++++++++++++++++++++++---- 4 files changed, 44 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/arch/sparc64/kernel/rtrap.S b/arch/sparc64/kernel/rtrap.S index b9b785fd8b46..16689b2930db 100644 --- a/arch/sparc64/kernel/rtrap.S +++ b/arch/sparc64/kernel/rtrap.S @@ -46,7 +46,7 @@ __handle_user_windows: wrpr %g0, RTRAP_PSTATE_IRQOFF, %pstate ldx [%g6 + TI_FLAGS], %l0 -1: andcc %l0, (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK), %g0 +1: andcc %l0, _TIF_SIGPENDING, %g0 be,pt %xcc, __handle_user_windows_continue nop mov %l5, %o1 @@ -86,7 +86,7 @@ __handle_perfctrs: wrpr %g0, RTRAP_PSTATE, %pstate wrpr %g0, RTRAP_PSTATE_IRQOFF, %pstate ldx [%g6 + TI_FLAGS], %l0 -1: andcc %l0, (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK), %g0 +1: andcc %l0, _TIF_SIGPENDING, %g0 be,pt %xcc, __handle_perfctrs_continue sethi %hi(TSTATE_PEF), %o0 @@ -195,7 +195,7 @@ __handle_preemption_continue: andcc %l1, %o0, %g0 andcc %l0, _TIF_NEED_RESCHED, %g0 bne,pn %xcc, __handle_preemption - andcc %l0, (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK), %g0 + andcc %l0, _TIF_SIGPENDING, %g0 bne,pn %xcc, __handle_signal __handle_signal_continue: ldub [%g6 + TI_WSAVED], %o2 diff --git a/arch/sparc64/kernel/signal.c b/arch/sparc64/kernel/signal.c index 2378482c2aab..6e4dc67d16af 100644 --- a/arch/sparc64/kernel/signal.c +++ b/arch/sparc64/kernel/signal.c @@ -247,7 +247,9 @@ static long _sigpause_common(old_sigset_t set) current->state = TASK_INTERRUPTIBLE; schedule(); - set_thread_flag(TIF_RESTORE_SIGMASK); + + set_restore_sigmask(); + return -ERESTARTNOHAND; } @@ -537,7 +539,7 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) } else restart_syscall = 0; - if (test_thread_flag(TIF_RESTORE_SIGMASK)) + if (current_thread_info()->status & TS_RESTORE_SIGMASK) oldset = ¤t->saved_sigmask; else oldset = ¤t->blocked; @@ -566,13 +568,12 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) syscall_restart(orig_i0, regs, &ka.sa); handle_signal(signr, &ka, &info, oldset, regs); - /* a signal was successfully delivered; the saved + /* A signal was successfully delivered; the saved * sigmask will have been stored in the signal frame, * and will be restored by sigreturn, so we can simply - * clear the TIF_RESTORE_SIGMASK flag. + * clear the TS_RESTORE_SIGMASK flag. */ - if (test_thread_flag(TIF_RESTORE_SIGMASK)) - clear_thread_flag(TIF_RESTORE_SIGMASK); + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; return; } if (restart_syscall && @@ -591,17 +592,17 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0) regs->tnpc -= 4; } - /* if there's no signal to deliver, we just put the saved sigmask + /* If there's no signal to deliver, we just put the saved sigmask * back */ - if (test_thread_flag(TIF_RESTORE_SIGMASK)) { - clear_thread_flag(TIF_RESTORE_SIGMASK); + if (current_thread_info()->status & TS_RESTORE_SIGMASK) { + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); } } void do_notify_resume(struct pt_regs *regs, unsigned long orig_i0, unsigned long thread_info_flags) { - if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK)) + if (thread_info_flags & _TIF_SIGPENDING) do_signal(regs, orig_i0); } diff --git a/arch/sparc64/kernel/signal32.c b/arch/sparc64/kernel/signal32.c index 3f19e9af3d1b..97cdd1bf4a10 100644 --- a/arch/sparc64/kernel/signal32.c +++ b/arch/sparc64/kernel/signal32.c @@ -788,13 +788,12 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs, syscall_restart32(orig_i0, regs, &ka.sa); handle_signal32(signr, &ka, &info, oldset, regs); - /* a signal was successfully delivered; the saved + /* A signal was successfully delivered; the saved * sigmask will have been stored in the signal frame, * and will be restored by sigreturn, so we can simply - * clear the TIF_RESTORE_SIGMASK flag. + * clear the TS_RESTORE_SIGMASK flag. */ - if (test_thread_flag(TIF_RESTORE_SIGMASK)) - clear_thread_flag(TIF_RESTORE_SIGMASK); + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; return; } if (restart_syscall && @@ -813,11 +812,11 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs, regs->tnpc -= 4; } - /* if there's no signal to deliver, we just put the saved sigmask + /* If there's no signal to deliver, we just put the saved sigmask * back */ - if (test_thread_flag(TIF_RESTORE_SIGMASK)) { - clear_thread_flag(TIF_RESTORE_SIGMASK); + if (current_thread_info()->status & TS_RESTORE_SIGMASK) { + current_thread_info()->status &= ~TS_RESTORE_SIGMASK; sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); } } diff --git a/include/asm-sparc64/thread_info.h b/include/asm-sparc64/thread_info.h index 71e42d1a80d9..e5873e385306 100644 --- a/include/asm-sparc64/thread_info.h +++ b/include/asm-sparc64/thread_info.h @@ -38,7 +38,7 @@ struct thread_info { struct task_struct *task; unsigned long flags; __u8 fpsaved[7]; - __u8 pad; + __u8 status; unsigned long ksp; /* D$ line 2 */ @@ -217,7 +217,7 @@ register struct thread_info *current_thread_info_reg asm("g6"); * nop */ #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ -#define TIF_RESTORE_SIGMASK 1 /* restore signal mask in do_signal() */ +/* flags bit 1 is available */ #define TIF_SIGPENDING 2 /* signal pending */ #define TIF_NEED_RESCHED 3 /* rescheduling necessary */ #define TIF_PERFCTR 4 /* performance counters active */ @@ -244,14 +244,34 @@ register struct thread_info *current_thread_info_reg asm("g6"); #define _TIF_32BIT (1<status |= TS_RESTORE_SIGMASK; + set_bit(TIF_SIGPENDING, &ti->flags); +} +#endif /* !__ASSEMBLY__ */ + #endif /* __KERNEL__ */ #endif /* _ASM_THREAD_INFO_H */ -- cgit From 392abe9cb36f04ccdda9c96938077b0d43f2b50b Mon Sep 17 00:00:00 2001 From: Philipp Zabel Date: Tue, 13 May 2008 14:03:40 +0200 Subject: [ALSA] ASoC: build fix for snd_soc_info_bool_ext I suspect that snd_ctl_boolean_mono should have been snd_ctl_boolean_mono_info instead. This fixes the build for magician. Signed-off-by: Philipp Zabel Acked-by: Mark Brown Signed-off-by: Takashi Iwai --- include/sound/soc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/sound/soc.h b/include/sound/soc.h index e6ea6f750941..d3c8c033dff8 100644 --- a/include/sound/soc.h +++ b/include/sound/soc.h @@ -238,7 +238,7 @@ int snd_soc_info_volsw(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo); int snd_soc_info_volsw_ext(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo); -#define snd_soc_info_bool_ext snd_ctl_boolean_mono +#define snd_soc_info_bool_ext snd_ctl_boolean_mono_info int snd_soc_get_volsw(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_value *ucontrol); int snd_soc_put_volsw(struct snd_kcontrol *kcontrol, -- cgit From 309e96cdf2f2c1a071102e8bdf828a3493e6e50a Mon Sep 17 00:00:00 2001 From: Jeff Dike Date: Mon, 12 May 2008 14:01:52 -0700 Subject: uml: remove unused header From: Adrian Bunk This patch removes the unused and broken (the normal asm/keyboard.h files no longer exists) include/asm-um/keyboard.h Signed-off-by: Adrian Bunk Signed-off-by: Jeff Dike Cc: WANG Cong Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-um/keyboard.h | 6 ------ 1 file changed, 6 deletions(-) delete mode 100644 include/asm-um/keyboard.h (limited to 'include') diff --git a/include/asm-um/keyboard.h b/include/asm-um/keyboard.h deleted file mode 100644 index ee2e2303d0e4..000000000000 --- a/include/asm-um/keyboard.h +++ /dev/null @@ -1,6 +0,0 @@ -#ifndef __UM_KEYBOARD_H -#define __UM_KEYBOARD_H - -#include "asm/arch/keyboard.h" - -#endif -- cgit From a7dfa9403bf3b03899d5ef5d10b0c5c3f74b0682 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Mon, 12 May 2008 14:01:56 -0700 Subject: uml: use PAGE_SIZE in linker scripts This patch includes page.h header into linker scripts that allow us to use PAGE_SIZE macro instead of numeric constant. To be able to include page.h into linker scripts page.h is needed for some modification - i.e. we need to use __ASSEMBLY__ and _AC macro [jdike@linux.intel.com - fixed conflict with as-layout.h] Signed-off-by: Cyrill Gorcunov Signed-off-by: Jeff Dike Cc: WANG Cong Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/include/as-layout.h | 14 +++++++------- arch/um/kernel/dyn.lds.S | 7 ++++--- arch/um/kernel/uml.lds.S | 7 ++++--- include/asm-um/page.h | 17 +++++++++++------ 4 files changed, 26 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/arch/um/include/as-layout.h b/arch/um/include/as-layout.h index cac542d8ff70..58e852dfb0ce 100644 --- a/arch/um/include/as-layout.h +++ b/arch/um/include/as-layout.h @@ -23,16 +23,16 @@ */ #ifdef __ASSEMBLY__ -#define _AC(X, Y) (Y) +#define _UML_AC(X, Y) (Y) #else -#define __AC(X, Y) (X (Y)) -#define _AC(X, Y) __AC(X, Y) +#define __UML_AC(X, Y) (X(Y)) +#define _UML_AC(X, Y) __UML_AC(X, Y) #endif -#define STUB_START _AC(, 0x100000) -#define STUB_CODE _AC((unsigned long), STUB_START) -#define STUB_DATA _AC((unsigned long), STUB_CODE + UM_KERN_PAGE_SIZE) -#define STUB_END _AC((unsigned long), STUB_DATA + UM_KERN_PAGE_SIZE) +#define STUB_START _UML_AC(, 0x100000) +#define STUB_CODE _UML_AC((unsigned long), STUB_START) +#define STUB_DATA _UML_AC((unsigned long), STUB_CODE + UM_KERN_PAGE_SIZE) +#define STUB_END _UML_AC((unsigned long), STUB_DATA + UM_KERN_PAGE_SIZE) #ifndef __ASSEMBLY__ diff --git a/arch/um/kernel/dyn.lds.S b/arch/um/kernel/dyn.lds.S index 26090b7f323e..9975e1ab44fb 100644 --- a/arch/um/kernel/dyn.lds.S +++ b/arch/um/kernel/dyn.lds.S @@ -1,4 +1,5 @@ #include +#include OUTPUT_FORMAT(ELF_FORMAT) OUTPUT_ARCH(ELF_ARCH) @@ -21,7 +22,7 @@ SECTIONS _einittext = .; } - . = ALIGN(4096); + . = ALIGN(PAGE_SIZE); /* Read-only sections, merged into text segment: */ .hash : { *(.hash) } @@ -68,9 +69,9 @@ SECTIONS /* .gnu.warning sections are handled specially by elf32.em. */ *(.gnu.warning) - . = ALIGN(4096); + . = ALIGN(PAGE_SIZE); } =0x90909090 - . = ALIGN(4096); + . = ALIGN(PAGE_SIZE); .syscall_stub : { __syscall_stub_start = .; *(.__syscall_stub*) diff --git a/arch/um/kernel/uml.lds.S b/arch/um/kernel/uml.lds.S index 5828c1d54505..11b835248b86 100644 --- a/arch/um/kernel/uml.lds.S +++ b/arch/um/kernel/uml.lds.S @@ -1,4 +1,5 @@ #include +#include OUTPUT_FORMAT(ELF_FORMAT) OUTPUT_ARCH(ELF_ARCH) @@ -26,7 +27,7 @@ SECTIONS INIT_TEXT _einittext = .; } - . = ALIGN(4096); + . = ALIGN(PAGE_SIZE); .text : { @@ -39,7 +40,7 @@ SECTIONS *(.gnu.linkonce.t*) } - . = ALIGN(4096); + . = ALIGN(PAGE_SIZE); .syscall_stub : { __syscall_stub_start = .; *(.__syscall_stub*) @@ -79,7 +80,7 @@ SECTIONS .sdata : { *(.sdata) } _edata = .; PROVIDE (edata = .); - . = ALIGN(0x1000); + . = ALIGN(PAGE_SIZE); .sbss : { __bss_start = .; diff --git a/include/asm-um/page.h b/include/asm-um/page.h index 381f96b1c825..916e1a61999f 100644 --- a/include/asm-um/page.h +++ b/include/asm-um/page.h @@ -7,16 +7,20 @@ #ifndef __UM_PAGE_H #define __UM_PAGE_H -struct page; - -#include -#include +#include /* PAGE_SHIFT determines the page size */ #define PAGE_SHIFT 12 -#define PAGE_SIZE (1UL << PAGE_SHIFT) +#define PAGE_SIZE (_AC(1, UL) << PAGE_SHIFT) #define PAGE_MASK (~(PAGE_SIZE-1)) +#ifndef __ASSEMBLY__ + +struct page; + +#include +#include + /* * These are used to make use of C type-checking.. */ @@ -120,4 +124,5 @@ extern struct page *arch_validate(struct page *page, gfp_t mask, int order); #include #include -#endif +#endif /* __ASSEMBLY__ */ +#endif /* __UM_PAGE_H */ -- cgit From 5d33e4d7fd9a52d2673e5c730eab81856e100a74 Mon Sep 17 00:00:00 2001 From: Jeff Dike Date: Mon, 12 May 2008 14:01:58 -0700 Subject: uml: random driver fixes The random driver would essentially hang if the host's /dev/random returned -EAGAIN. There was a test of need_resched followed by a schedule inside the loop, but that didn't help and it's the wrong way to work anyway. The right way is to ask for an interrupt when there is input available from the host and handle it then rather than polling. Now, when the host's /dev/random returns -EAGAIN, the driver asks for a wakeup when there's randomness available again and sleeps. The interrupt routine just wakes up whatever processes are sleeping on host_read_wait. There is an atomic_t, host_sleep_count, which counts the number of processes waiting for randomness. When this reaches zero, the interrupt is disabled. An added complication is that async I/O notification was only recently added to /dev/random (by me), so essentially all hosts will lack it. So, we use the sigio workaround here, which is to have a separate thread poll on the descriptor and send an interrupt when there is input on it. This mechanism is activated when a process gets -EAGAIN (activating this multiple times is harmless, if a bit wasteful) and deactivated by the last process still waiting. The module name was changed from "random" to "hw_random" in order for udev to recognize it. The sigio workaround needed some changes. sigio_broken was added for cases when we know that async notification doesn't work. This is now called from maybe_sigio_broken, which deals with pts devices. Signed-off-by: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/random.c | 41 +++++++++++++++++++++++++++++++++++++---- arch/um/include/os.h | 1 + arch/um/include/process.h | 16 ++++------------ arch/um/os-Linux/sigio.c | 35 +++++++++++++++++++++-------------- include/asm-um/irq.h | 3 ++- 5 files changed, 65 insertions(+), 31 deletions(-) (limited to 'include') diff --git a/arch/um/drivers/random.c b/arch/um/drivers/random.c index 71f0959c1535..f92b7c81eb00 100644 --- a/arch/um/drivers/random.c +++ b/arch/um/drivers/random.c @@ -8,16 +8,18 @@ #include #include #include +#include #include #include #include +#include "irq_kern.h" #include "os.h" /* * core module and version information */ #define RNG_VERSION "1.0.0" -#define RNG_MODULE_NAME "random" +#define RNG_MODULE_NAME "hw_random" #define RNG_MISCDEV_MINOR 183 /* official */ @@ -26,6 +28,7 @@ * protects against a module being loaded twice at the same time. */ static int random_fd = -1; +static DECLARE_WAIT_QUEUE_HEAD(host_read_wait); static int rng_dev_open (struct inode *inode, struct file *filp) { @@ -38,6 +41,8 @@ static int rng_dev_open (struct inode *inode, struct file *filp) return 0; } +static atomic_t host_sleep_count = ATOMIC_INIT(0); + static ssize_t rng_dev_read (struct file *filp, char __user *buf, size_t size, loff_t * offp) { @@ -60,11 +65,26 @@ static ssize_t rng_dev_read (struct file *filp, char __user *buf, size_t size, } } else if(n == -EAGAIN){ + DECLARE_WAITQUEUE(wait, current); + if (filp->f_flags & O_NONBLOCK) return ret ? : -EAGAIN; - if(need_resched()) - schedule_timeout_interruptible(1); + atomic_inc(&host_sleep_count); + reactivate_fd(random_fd, RANDOM_IRQ); + add_sigio_fd(random_fd); + + add_wait_queue(&host_read_wait, &wait); + set_task_state(current, TASK_INTERRUPTIBLE); + + schedule(); + set_task_state(current, TASK_RUNNING); + remove_wait_queue(&host_read_wait, &wait); + + if (atomic_dec_and_test(&host_sleep_count)) { + ignore_sigio_fd(random_fd); + deactivate_fd(random_fd, RANDOM_IRQ); + } } else return n; if (signal_pending (current)) @@ -86,6 +106,13 @@ static struct miscdevice rng_miscdev = { &rng_chrdev_ops, }; +static irqreturn_t random_interrupt(int irq, void *data) +{ + wake_up(&host_read_wait); + + return IRQ_HANDLED; +} + /* * rng_init - initialize RNG module */ @@ -99,10 +126,14 @@ static int __init rng_init (void) random_fd = err; - err = os_set_fd_block(random_fd, 0); + err = um_request_irq(RANDOM_IRQ, random_fd, IRQ_READ, random_interrupt, + IRQF_DISABLED | IRQF_SAMPLE_RANDOM, "random", + NULL); if(err) goto err_out_cleanup_hw; + sigio_broken(random_fd, 1); + err = misc_register (&rng_miscdev); if (err) { printk (KERN_ERR RNG_MODULE_NAME ": misc device register failed\n"); @@ -113,6 +144,7 @@ static int __init rng_init (void) return err; err_out_cleanup_hw: + os_close_file(random_fd); random_fd = -1; goto out; } @@ -122,6 +154,7 @@ static int __init rng_init (void) */ static void __exit rng_cleanup (void) { + os_close_file(random_fd); misc_deregister (&rng_miscdev); } diff --git a/arch/um/include/os.h b/arch/um/include/os.h index 32c799e3a495..e2716ac8889a 100644 --- a/arch/um/include/os.h +++ b/arch/um/include/os.h @@ -290,6 +290,7 @@ extern void os_set_ioignore(void); extern int add_sigio_fd(int fd); extern int ignore_sigio_fd(int fd); extern void maybe_sigio_broken(int fd, int read); +extern void sigio_broken(int fd, int read); /* sys-x86_64/prctl.c */ extern int os_arch_prctl(int pid, int code, unsigned long *addr); diff --git a/arch/um/include/process.h b/arch/um/include/process.h index 5af9157ff54f..838b4802ce53 100644 --- a/arch/um/include/process.h +++ b/arch/um/include/process.h @@ -1,5 +1,5 @@ /* - * Copyright (C) 2000, 2001, 2002 Jeff Dike (jdike@karaya.com) + * Copyright (C) 2000 - 2008 Jeff Dike (jdike@{addtoit,linux.intel}.com) * Licensed under the GPL */ @@ -8,18 +8,10 @@ #include +/* Copied from linux/compiler-gcc.h since we can't include it directly */ +#define barrier() __asm__ __volatile__("": : :"memory") + extern void sig_handler(int sig, struct sigcontext sc); extern void alarm_handler(int sig, struct sigcontext sc); #endif - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/os-Linux/sigio.c b/arch/um/os-Linux/sigio.c index 0578481983c4..eb8f2e4be192 100644 --- a/arch/um/os-Linux/sigio.c +++ b/arch/um/os-Linux/sigio.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Copyright (C) 2002 - 2008 Jeff Dike (jdike@{addtoit,linux.intel}.com) * Licensed under the GPL */ @@ -15,6 +15,7 @@ #include "kern_util.h" #include "init.h" #include "os.h" +#include "process.h" #include "sigio.h" #include "um_malloc.h" #include "user.h" @@ -338,20 +339,10 @@ out_close1: close(l_write_sigio_fds[1]); } -/* Changed during early boot */ -static int pty_output_sigio = 0; -static int pty_close_sigio = 0; - -void maybe_sigio_broken(int fd, int read) +void sigio_broken(int fd, int read) { int err; - if (!isatty(fd)) - return; - - if ((read || pty_output_sigio) && (!read || pty_close_sigio)) - return; - write_sigio_workaround(); sigio_lock(); @@ -370,6 +361,21 @@ out: sigio_unlock(); } +/* Changed during early boot */ +static int pty_output_sigio; +static int pty_close_sigio; + +void maybe_sigio_broken(int fd, int read) +{ + if (!isatty(fd)) + return; + + if ((read || pty_output_sigio) && (!read || pty_close_sigio)) + return; + + sigio_broken(fd, read); +} + static void sigio_cleanup(void) { if (write_sigio_pid == -1) @@ -383,7 +389,7 @@ static void sigio_cleanup(void) __uml_exitcall(sigio_cleanup); /* Used as a flag during SIGIO testing early in boot */ -static volatile int got_sigio = 0; +static int got_sigio; static void __init handler(int sig) { @@ -498,7 +504,8 @@ static void tty_output(int master, int slave) if (errno != EAGAIN) printk(UM_KERN_ERR "tty_output : write failed, errno = %d\n", errno); - while (((n = read(slave, buf, sizeof(buf))) > 0) && !got_sigio) + while (((n = read(slave, buf, sizeof(buf))) > 0) && + !({ barrier(); got_sigio; })) ; if (got_sigio) { diff --git a/include/asm-um/irq.h b/include/asm-um/irq.h index de389a477cdd..4a2037f8204b 100644 --- a/include/asm-um/irq.h +++ b/include/asm-um/irq.h @@ -15,8 +15,9 @@ #define SIGIO_WRITE_IRQ 11 #define TELNETD_IRQ 12 #define XTERM_IRQ 13 +#define RANDOM_IRQ 14 -#define LAST_IRQ XTERM_IRQ +#define LAST_IRQ RANDOM_IRQ #define NR_IRQS (LAST_IRQ + 1) #endif -- cgit From f4ed0deae8983591264d0e194e168ef65f4775f5 Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Mon, 12 May 2008 14:02:29 -0700 Subject: cpumask: remove bitmap_scnprintf_len and cpumask_scnprintf_len They aren't used. They were briefly used as part of some other patches to provide an alternative format for displaying some /proc and /sys cpumasks. They probably should have been removed when those other patches were dropped, in favor of a different solution. Signed-off-by: Paul Jackson Cc: "Mike Travis" Cc: "Bert Wesarg" Cc: Alexey Dobriyan Cc: WANG Cong Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/bitmap.h | 1 - include/linux/cpumask.h | 7 ------- lib/bitmap.c | 16 ---------------- 3 files changed, 24 deletions(-) (limited to 'include') diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h index 43b406def35f..1abfe664c444 100644 --- a/include/linux/bitmap.h +++ b/include/linux/bitmap.h @@ -110,7 +110,6 @@ extern int __bitmap_weight(const unsigned long *bitmap, int bits); extern int bitmap_scnprintf(char *buf, unsigned int len, const unsigned long *src, int nbits); -extern int bitmap_scnprintf_len(unsigned int len); extern int __bitmap_parse(const char *buf, unsigned int buflen, int is_user, unsigned long *dst, int nbits); extern int bitmap_parse_user(const char __user *ubuf, unsigned int ulen, diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h index 9650806fe2ea..5df3db58fcc6 100644 --- a/include/linux/cpumask.h +++ b/include/linux/cpumask.h @@ -289,13 +289,6 @@ static inline int __cpumask_scnprintf(char *buf, int len, return bitmap_scnprintf(buf, len, srcp->bits, nbits); } -#define cpumask_scnprintf_len(len) \ - __cpumask_scnprintf_len((len)) -static inline int __cpumask_scnprintf_len(int len) -{ - return bitmap_scnprintf_len(len); -} - #define cpumask_parse_user(ubuf, ulen, dst) \ __cpumask_parse_user((ubuf), (ulen), &(dst), NR_CPUS) static inline int __cpumask_parse_user(const char __user *buf, int len, diff --git a/lib/bitmap.c b/lib/bitmap.c index c4cb48f77f0c..482df94ea21e 100644 --- a/lib/bitmap.c +++ b/lib/bitmap.c @@ -315,22 +315,6 @@ int bitmap_scnprintf(char *buf, unsigned int buflen, } EXPORT_SYMBOL(bitmap_scnprintf); -/** - * bitmap_scnprintf_len - return buffer length needed to convert - * bitmap to an ASCII hex string. - * @len: number of bits to be converted - */ -int bitmap_scnprintf_len(unsigned int len) -{ - /* we need 9 chars per word for 32 bit words (8 hexdigits + sep/null) */ - int bitslen = ALIGN(len, CHUNKSZ); - int wordlen = CHUNKSZ / 4; - int buflen = (bitslen / wordlen) * (wordlen + 1) * sizeof(char); - - return buflen; -} -EXPORT_SYMBOL(bitmap_scnprintf_len); - /** * __bitmap_parse - convert an ASCII hex string into a bitmap. * @buf: pointer to buffer containing string. -- cgit From 4cd1a8fc3d3cd740416b14ece2693dbb5d065eaf Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Mon, 12 May 2008 14:02:31 -0700 Subject: memcg: fix possible panic when CONFIG_MM_OWNER=y When mm destruction happens, we should pass mm_update_next_owner() the old mm. But unfortunately new mm is passed in exec_mmap(). Thus, kernel panic is possible when a multi-threaded process uses exec(). Also, the owner member comment description is wrong. mm->owner does not necessarily point to the thread group leader. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: KOSAKI Motohiro Acked-by: Balbir Singh Cc: "Paul Menage" Cc: "KAMEZAWA Hiroyuki" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 2 +- include/linux/mm_types.h | 13 +++++++++++-- 2 files changed, 12 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/fs/exec.c b/fs/exec.c index aeaa9791d8be..1f8a24aa1f8b 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -736,7 +736,7 @@ static int exec_mmap(struct mm_struct *mm) tsk->active_mm = mm; activate_mm(active_mm, mm); task_unlock(tsk); - mm_update_next_owner(mm); + mm_update_next_owner(old_mm); arch_pick_mmap_layout(mm); if (old_mm) { up_read(&old_mm->mmap_sem); diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index eb7c16cc9559..02a27ae78539 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -226,8 +226,17 @@ struct mm_struct { rwlock_t ioctx_list_lock; /* aio lock */ struct kioctx *ioctx_list; #ifdef CONFIG_MM_OWNER - struct task_struct *owner; /* The thread group leader that */ - /* owns the mm_struct. */ + /* + * "owner" points to a task that is regarded as the canonical + * user/owner of this mm. All of the following must be true in + * order for it to be changed: + * + * current == mm->owner + * current->mm != mm + * new_owner->mm == mm + * new_owner->alloc_lock is held + */ + struct task_struct *owner; #endif #ifdef CONFIG_PROC_FS -- cgit From 78bb6cb9a890d3d50ca3b02fce9223d3e734ab9b Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Mon, 12 May 2008 14:02:32 -0700 Subject: fuse: add flag to turn on big writes Prior to 2.6.26 fuse only supported single page write requests. In theory all fuse filesystem should be able support bigger than 4k writes, as there's nothing in the API to prevent it. Unfortunately there's a known case in NTFS-3G where big writes cause filesystem corruption. There could also be other filesystems, where the lack of testing with big write requests would result in bugs. To prevent such problems on a kernel upgrade, disable big writes by default, but let filesystems set a flag to turn it on. Signed-off-by: Miklos Szeredi Cc: Szabolcs Szakacsits Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fuse/file.c | 2 ++ fs/fuse/fuse_i.h | 3 +++ fs/fuse/inode.c | 5 ++++- include/linux/fuse.h | 1 + 4 files changed, 10 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index f28cf8b46f80..8092f0d9fd1f 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -804,6 +804,8 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, if (offset == PAGE_CACHE_SIZE) offset = 0; + if (!fc->big_writes) + break; } while (iov_iter_count(ii) && count < fc->max_write && req->num_pages < FUSE_MAX_PAGES_PER_REQ && offset == 0); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index dadffa21a206..bae948657c4f 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -404,6 +404,9 @@ struct fuse_conn { /** Is bmap not implemented by fs? */ unsigned no_bmap : 1; + /** Do multi-page cached writes */ + unsigned big_writes : 1; + /** The number of requests waiting for completion */ atomic_t num_waiting; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 79b615873838..fb77e0962132 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -576,6 +576,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) fc->no_lock = 1; if (arg->flags & FUSE_ATOMIC_O_TRUNC) fc->atomic_o_trunc = 1; + if (arg->flags & FUSE_BIG_WRITES) + fc->big_writes = 1; } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; fc->no_lock = 1; @@ -599,7 +601,8 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) arg->major = FUSE_KERNEL_VERSION; arg->minor = FUSE_KERNEL_MINOR_VERSION; arg->max_readahead = fc->bdi.ra_pages * PAGE_CACHE_SIZE; - arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS | FUSE_ATOMIC_O_TRUNC; + arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS | FUSE_ATOMIC_O_TRUNC | + FUSE_BIG_WRITES; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); diff --git a/include/linux/fuse.h b/include/linux/fuse.h index 5c86f1196c3a..d48282197696 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -109,6 +109,7 @@ struct fuse_file_lock { #define FUSE_POSIX_LOCKS (1 << 1) #define FUSE_FILE_OPS (1 << 2) #define FUSE_ATOMIC_O_TRUNC (1 << 3) +#define FUSE_BIG_WRITES (1 << 5) /** * Release flags -- cgit From 332223831e86b2e17b48b4afafad07d8e3b73861 Mon Sep 17 00:00:00 2001 From: Graf Yang Date: Tue, 13 May 2008 23:25:57 -0700 Subject: irda: Fix a misalign access issue. (v2) Replace u16ho with put/get_unaligned functions Signed-off-by: Graf Yang Signed-off-by: Bryan Wu Signed-off-by: David S. Miller --- include/net/irda/discovery.h | 3 --- net/irda/discovery.c | 8 +++++--- net/irda/irlmp.c | 5 +++-- net/irda/irnet/irnet_irda.c | 4 ++-- 4 files changed, 10 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/include/net/irda/discovery.h b/include/net/irda/discovery.h index e4efad1f9eff..0ce93398720d 100644 --- a/include/net/irda/discovery.h +++ b/include/net/irda/discovery.h @@ -57,9 +57,6 @@ typedef union { __u8 byte[2]; } __u16_host_order; -/* Same purpose, different application */ -#define u16ho(array) (* ((__u16 *) array)) - /* Types of discovery */ typedef enum { DISCOVERY_LOG, /* What's in our discovery log */ diff --git a/net/irda/discovery.c b/net/irda/discovery.c index bfacef8b76f4..a6f99b5a1499 100644 --- a/net/irda/discovery.c +++ b/net/irda/discovery.c @@ -40,6 +40,8 @@ #include +#include + /* * Function irlmp_add_discovery (cachelog, discovery) * @@ -87,7 +89,7 @@ void irlmp_add_discovery(hashbin_t *cachelog, discovery_t *new) */ hashbin_remove_this(cachelog, (irda_queue_t *) node); /* Check if hints bits are unchanged */ - if(u16ho(node->data.hints) == u16ho(new->data.hints)) + if (get_unaligned((__u16 *)node->data.hints) == get_unaligned((__u16 *)new->data.hints)) /* Set time of first discovery for this node */ new->firststamp = node->firststamp; kfree(node); @@ -281,9 +283,9 @@ struct irda_device_info *irlmp_copy_discoveries(hashbin_t *log, int *pn, /* Mask out the ones we don't want : * We want to match the discovery mask, and to get only * the most recent one (unless we want old ones) */ - if ((u16ho(discovery->data.hints) & mask) && + if ((get_unaligned((__u16 *)discovery->data.hints) & mask) && ((old_entries) || - ((jiffies - discovery->firststamp) < j_timeout)) ) { + ((jiffies - discovery->firststamp) < j_timeout))) { /* Create buffer as needed. * As this function get called a lot and most time * we don't have anything to put in the log (we are diff --git a/net/irda/irlmp.c b/net/irda/irlmp.c index 1f81f8e7c61d..7bf5b913828b 100644 --- a/net/irda/irlmp.c +++ b/net/irda/irlmp.c @@ -1062,7 +1062,8 @@ void irlmp_discovery_expiry(discinfo_t *expiries, int number) for(i = 0; i < number; i++) { /* Check if we should notify client */ if ((client->expir_callback) && - (client->hint_mask.word & u16ho(expiries[i].hints) + (client->hint_mask.word & + get_unaligned((__u16 *)expiries[i].hints) & 0x7f7f) ) client->expir_callback(&(expiries[i]), EXPIRY_TIMEOUT, @@ -1086,7 +1087,7 @@ discovery_t *irlmp_get_discovery_response(void) IRDA_ASSERT(irlmp != NULL, return NULL;); - u16ho(irlmp->discovery_rsp.data.hints) = irlmp->hints.word; + put_unaligned(irlmp->hints.word, (__u16 *)irlmp->discovery_rsp.data.hints); /* * Set character set for device name (we use ASCII), and diff --git a/net/irda/irnet/irnet_irda.c b/net/irda/irnet/irnet_irda.c index 75497e55927d..a3ec0026cdb2 100644 --- a/net/irda/irnet/irnet_irda.c +++ b/net/irda/irnet/irnet_irda.c @@ -1673,7 +1673,7 @@ irnet_discovery_indication(discinfo_t * discovery, /* Notify the control channel */ irnet_post_event(NULL, IRNET_DISCOVER, discovery->saddr, discovery->daddr, discovery->info, - u16ho(discovery->hints)); + get_unaligned((__u16 *)discovery->hints)); DEXIT(IRDA_OCB_TRACE, "\n"); } @@ -1704,7 +1704,7 @@ irnet_expiry_indication(discinfo_t * expiry, /* Notify the control channel */ irnet_post_event(NULL, IRNET_EXPIRE, expiry->saddr, expiry->daddr, expiry->info, - u16ho(expiry->hints)); + get_unaligned((__u16 *)expiry->hints)); DEXIT(IRDA_OCB_TRACE, "\n"); } -- cgit From d23039eec77473124c9635c01378314f196f2211 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Fri, 25 Apr 2008 19:23:16 -0700 Subject: USB: add association.h This will be used by the wireless usb code, as well as potentially other USB code. Originally based on some .c code written by Inaky Perez-Gonzalez Cc: Inaky Perez-Gonzalez Cc: David Brownell Signed-off-by: Greg Kroah-Hartman --- include/linux/usb/association.h | 150 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 150 insertions(+) create mode 100644 include/linux/usb/association.h (limited to 'include') diff --git a/include/linux/usb/association.h b/include/linux/usb/association.h new file mode 100644 index 000000000000..07c5e3cf5898 --- /dev/null +++ b/include/linux/usb/association.h @@ -0,0 +1,150 @@ +/* + * Wireless USB - Cable Based Association + * + * Copyright (C) 2006 Intel Corporation + * Inaky Perez-Gonzalez + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version + * 2 as published by the Free Software Foundation. + * + */ +#ifndef __LINUX_USB_ASSOCIATION_H +#define __LINUX_USB_ASSOCIATION_H + + +/* + * Association attributes + * + * Association Models Supplement to WUSB 1.0 T[3-1] + * + * Each field in the structures has it's ID, it's length and then the + * value. This is the actual definition of the field's ID and its + * length. + */ +struct wusb_am_attr { + __u8 id; + __u8 len; +}; + +/* Different fields defined by the spec */ +#define WUSB_AR_AssociationTypeId { .id = 0x0000, .len = 2 } +#define WUSB_AR_AssociationSubTypeId { .id = 0x0001, .len = 2 } +#define WUSB_AR_Length { .id = 0x0002, .len = 4 } +#define WUSB_AR_AssociationStatus { .id = 0x0004, .len = 4 } +#define WUSB_AR_LangID { .id = 0x0008, .len = 2 } +#define WUSB_AR_DeviceFriendlyName { .id = 0x000b, .len = 64 } /* max */ +#define WUSB_AR_HostFriendlyName { .id = 0x000c, .len = 64 } /* max */ +#define WUSB_AR_CHID { .id = 0x1000, .len = 16 } +#define WUSB_AR_CDID { .id = 0x1001, .len = 16 } +#define WUSB_AR_ConnectionContext { .id = 0x1002, .len = 48 } +#define WUSB_AR_BandGroups { .id = 0x1004, .len = 2 } + +/* CBAF Control Requests (AMS1.0[T4-1] */ +enum { + CBAF_REQ_GET_ASSOCIATION_INFORMATION = 0x01, + CBAF_REQ_GET_ASSOCIATION_REQUEST, + CBAF_REQ_SET_ASSOCIATION_RESPONSE +}; + +/* + * CBAF USB-interface defitions + * + * No altsettings, one optional interrupt endpoint. + */ +enum { + CBAF_IFACECLASS = 0xef, + CBAF_IFACESUBCLASS = 0x03, + CBAF_IFACEPROTOCOL = 0x01, +}; + +/* Association Information (AMS1.0[T4-3]) */ +struct wusb_cbaf_assoc_info { + __le16 Length; + __u8 NumAssociationRequests; + __le16 Flags; + __u8 AssociationRequestsArray[]; +} __attribute__((packed)); + +/* Association Request (AMS1.0[T4-4]) */ +struct wusb_cbaf_assoc_request { + __u8 AssociationDataIndex; + __u8 Reserved; + __le16 AssociationTypeId; + __le16 AssociationSubTypeId; + __le32 AssociationTypeInfoSize; +} __attribute__((packed)); + +enum { + AR_TYPE_WUSB = 0x0001, + AR_TYPE_WUSB_RETRIEVE_HOST_INFO = 0x0000, + AR_TYPE_WUSB_ASSOCIATE = 0x0001, +}; + +/* Association Attribute header (AMS1.0[3.8]) */ +struct wusb_cbaf_attr_hdr { + __le16 id; + __le16 len; +} __attribute__((packed)); + +/* Host Info (AMS1.0[T4-7]) (yeah, more headers and fields...) */ +struct wusb_cbaf_host_info { + struct wusb_cbaf_attr_hdr AssociationTypeId_hdr; + __le16 AssociationTypeId; + struct wusb_cbaf_attr_hdr AssociationSubTypeId_hdr; + __le16 AssociationSubTypeId; + struct wusb_cbaf_attr_hdr CHID_hdr; + struct wusb_ckhdid CHID; + struct wusb_cbaf_attr_hdr LangID_hdr; + __le16 LangID; + struct wusb_cbaf_attr_hdr HostFriendlyName_hdr; + __u8 HostFriendlyName[]; +} __attribute__((packed)); + +/* Device Info (AMS1.0[T4-8]) + * + * I still don't get this tag'n'header stuff for each goddamn + * field... + */ +struct wusb_cbaf_device_info { + struct wusb_cbaf_attr_hdr Length_hdr; + __le32 Length; + struct wusb_cbaf_attr_hdr CDID_hdr; + struct wusb_ckhdid CDID; + struct wusb_cbaf_attr_hdr BandGroups_hdr; + __le16 BandGroups; + struct wusb_cbaf_attr_hdr LangID_hdr; + __le16 LangID; + struct wusb_cbaf_attr_hdr DeviceFriendlyName_hdr; + __u8 DeviceFriendlyName[]; +} __attribute__((packed)); + +/* Connection Context; CC_DATA - Success case (AMS1.0[T4-9]) */ +struct wusb_cbaf_cc_data { + struct wusb_cbaf_attr_hdr AssociationTypeId_hdr; + __le16 AssociationTypeId; + struct wusb_cbaf_attr_hdr AssociationSubTypeId_hdr; + __le16 AssociationSubTypeId; + struct wusb_cbaf_attr_hdr Length_hdr; + __le32 Length; + struct wusb_cbaf_attr_hdr ConnectionContext_hdr; + struct wusb_ckhdid CHID; + struct wusb_ckhdid CDID; + struct wusb_ckhdid CK; + struct wusb_cbaf_attr_hdr BandGroups_hdr; + __le16 BandGroups; +} __attribute__((packed)); + +/* CC_DATA - Failure case (AMS1.0[T4-10]) */ +struct wusb_cbaf_cc_data_fail { + struct wusb_cbaf_attr_hdr AssociationTypeId_hdr; + __le16 AssociationTypeId; + struct wusb_cbaf_attr_hdr AssociationSubTypeId_hdr; + __le16 AssociationSubTypeId; + struct wusb_cbaf_attr_hdr Length_hdr; + __le16 Length; + struct wusb_cbaf_attr_hdr AssociationStatus_hdr; + __u32 AssociationStatus; +} __attribute__((packed)); + +#endif /* __LINUX_USB_ASSOCIATION_H */ -- cgit From 73f10281ea96d7e8b4fc1c5d755a7c8eb484155b Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 14 May 2008 06:35:11 +0200 Subject: read_barrier_depends arch fixlets read_barrie_depends has always been a noop (not a compiler barrier) on all architectures except SMP alpha. This brings UP alpha and frv into line with all other architectures, and fixes incorrect documentation. Signed-off-by: Nick Piggin Acked-by: Paul E. McKenney Signed-off-by: Linus Torvalds --- Documentation/memory-barriers.txt | 12 +++++++++++- include/asm-alpha/barrier.h | 2 +- include/asm-frv/system.h | 2 +- 3 files changed, 13 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt index e5a819a4f0c9..f5b7127f54ac 100644 --- a/Documentation/memory-barriers.txt +++ b/Documentation/memory-barriers.txt @@ -994,7 +994,17 @@ The Linux kernel has eight basic CPU memory barriers: DATA DEPENDENCY read_barrier_depends() smp_read_barrier_depends() -All CPU memory barriers unconditionally imply compiler barriers. +All memory barriers except the data dependency barriers imply a compiler +barrier. Data dependencies do not impose any additional compiler ordering. + +Aside: In the case of data dependencies, the compiler would be expected to +issue the loads in the correct order (eg. `a[b]` would have to load the value +of b before loading a[b]), however there is no guarantee in the C specification +that the compiler may not speculate the value of b (eg. is equal to 1) and load +a before b (eg. tmp = a[1]; if (b != 1) tmp = a[b]; ). There is also the +problem of a compiler reloading b after having loaded a[b], thus having a newer +copy of b than a[b]. A consensus has not yet been reached about these problems, +however the ACCESS_ONCE macro is a good place to start looking. SMP memory barriers are reduced to compiler barriers on uniprocessor compiled systems because it is assumed that a CPU will appear to be self-consistent, diff --git a/include/asm-alpha/barrier.h b/include/asm-alpha/barrier.h index 384dc08d6f53..ac78eba909bc 100644 --- a/include/asm-alpha/barrier.h +++ b/include/asm-alpha/barrier.h @@ -24,7 +24,7 @@ __asm__ __volatile__("mb": : :"memory") #define smp_mb() barrier() #define smp_rmb() barrier() #define smp_wmb() barrier() -#define smp_read_barrier_depends() barrier() +#define smp_read_barrier_depends() do { } while (0) #endif #define set_mb(var, value) \ diff --git a/include/asm-frv/system.h b/include/asm-frv/system.h index cb307f8a6b48..d3a12a9079f7 100644 --- a/include/asm-frv/system.h +++ b/include/asm-frv/system.h @@ -179,7 +179,7 @@ do { \ #define mb() asm volatile ("membar" : : :"memory") #define rmb() asm volatile ("membar" : : :"memory") #define wmb() asm volatile ("membar" : : :"memory") -#define read_barrier_depends() barrier() +#define read_barrier_depends() do { } while (0) #ifdef CONFIG_SMP #define smp_mb() mb() -- cgit From 362a61ad61199e19a61b8e432015e2586b288f5b Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 14 May 2008 06:37:36 +0200 Subject: fix SMP data race in pagetable setup vs walking There is a possible data race in the page table walking code. After the split ptlock patches, it actually seems to have been introduced to the core code, but even before that I think it would have impacted some architectures (powerpc and sparc64, at least, walk the page tables without taking locks eg. see find_linux_pte()). The race is as follows: The pte page is allocated, zeroed, and its struct page gets its spinlock initialized. The mm-wide ptl is then taken, and then the pte page is inserted into the pagetables. At this point, the spinlock is not guaranteed to have ordered the previous stores to initialize the pte page with the subsequent store to put it in the page tables. So another Linux page table walker might be walking down (without any locks, because we have split-leaf-ptls), and find that new pte we've inserted. It might try to take the spinlock before the store from the other CPU initializes it. And subsequently it might read a pte_t out before stores from the other CPU have cleared the memory. There are also similar races in higher levels of the page tables. They obviously don't involve the spinlock, but could see uninitialized memory. Arch code and hardware pagetable walkers that walk the pagetables without locks could see similar uninitialized memory problems, regardless of whether split ptes are enabled or not. I prefer to put the barriers in core code, because that's where the higher level logic happens, but the page table accessors are per-arch, and open-coding them everywhere I don't think is an option. I'll put the read-side barriers in alpha arch code for now (other architectures perform data-dependent loads in order). Signed-off-by: Nick Piggin Signed-off-by: Linus Torvalds --- include/asm-alpha/pgtable.h | 21 +++++++++++++++++++-- mm/memory.c | 21 +++++++++++++++++++++ 2 files changed, 40 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/asm-alpha/pgtable.h b/include/asm-alpha/pgtable.h index 05ce5fba43e3..3f0c59f6d8aa 100644 --- a/include/asm-alpha/pgtable.h +++ b/include/asm-alpha/pgtable.h @@ -287,17 +287,34 @@ extern inline pte_t pte_mkspecial(pte_t pte) { return pte; } #define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1)) #define pgd_offset(mm, address) ((mm)->pgd+pgd_index(address)) +/* + * The smp_read_barrier_depends() in the following functions are required to + * order the load of *dir (the pointer in the top level page table) with any + * subsequent load of the returned pmd_t *ret (ret is data dependent on *dir). + * + * If this ordering is not enforced, the CPU might load an older value of + * *ret, which may be uninitialized data. See mm/memory.c:__pte_alloc for + * more details. + * + * Note that we never change the mm->pgd pointer after the task is running, so + * pgd_offset does not require such a barrier. + */ + /* Find an entry in the second-level page table.. */ extern inline pmd_t * pmd_offset(pgd_t * dir, unsigned long address) { - return (pmd_t *) pgd_page_vaddr(*dir) + ((address >> PMD_SHIFT) & (PTRS_PER_PAGE - 1)); + pmd_t *ret = (pmd_t *) pgd_page_vaddr(*dir) + ((address >> PMD_SHIFT) & (PTRS_PER_PAGE - 1)); + smp_read_barrier_depends(); /* see above */ + return ret; } /* Find an entry in the third-level page table.. */ extern inline pte_t * pte_offset_kernel(pmd_t * dir, unsigned long address) { - return (pte_t *) pmd_page_vaddr(*dir) + pte_t *ret = (pte_t *) pmd_page_vaddr(*dir) + ((address >> PAGE_SHIFT) & (PTRS_PER_PAGE - 1)); + smp_read_barrier_depends(); /* see above */ + return ret; } #define pte_offset_map(dir,addr) pte_offset_kernel((dir),(addr)) diff --git a/mm/memory.c b/mm/memory.c index 48c122d42ed7..fb5608a120ed 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -311,6 +311,21 @@ int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) if (!new) return -ENOMEM; + /* + * Ensure all pte setup (eg. pte page lock and page clearing) are + * visible before the pte is made visible to other CPUs by being + * put into page tables. + * + * The other side of the story is the pointer chasing in the page + * table walking code (when walking the page table without locking; + * ie. most of the time). Fortunately, these data accesses consist + * of a chain of data-dependent loads, meaning most CPUs (alpha + * being the notable exception) will already guarantee loads are + * seen in-order. See the alpha page table accessors for the + * smp_read_barrier_depends() barriers in page table walking code. + */ + smp_wmb(); /* Could be smp_wmb__xxx(before|after)_spin_lock */ + spin_lock(&mm->page_table_lock); if (!pmd_present(*pmd)) { /* Has another populated it ? */ mm->nr_ptes++; @@ -329,6 +344,8 @@ int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) if (!new) return -ENOMEM; + smp_wmb(); /* See comment in __pte_alloc */ + spin_lock(&init_mm.page_table_lock); if (!pmd_present(*pmd)) { /* Has another populated it ? */ pmd_populate_kernel(&init_mm, pmd, new); @@ -2619,6 +2636,8 @@ int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) if (!new) return -ENOMEM; + smp_wmb(); /* See comment in __pte_alloc */ + spin_lock(&mm->page_table_lock); if (pgd_present(*pgd)) /* Another has populated it */ pud_free(mm, new); @@ -2640,6 +2659,8 @@ int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) if (!new) return -ENOMEM; + smp_wmb(); /* See comment in __pte_alloc */ + spin_lock(&mm->page_table_lock); #ifndef __ARCH_HAS_4LEVEL_HACK if (pud_present(*pud)) /* Another has populated it */ -- cgit From 30f2f0eb4bd2c43d10a8b0d872c6e5ad8f31c9a0 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Tue, 6 May 2008 22:31:33 +0200 Subject: block: do_mounts - accept root= Some devices, like md, may create partitions only at first access, so allow root= to be set to a valid non-existant partition of an existing disk. This applies only to non-initramfs root mounting. This fixes a regression from 2.6.24 which did allow this to happen and broke some users machines :( Acked-by: Neil Brown Tested-by: Joao Luis Meloni Assirati Cc: stable Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- block/genhd.c | 9 ++++++--- include/linux/genhd.h | 4 ++-- init/do_mounts.c | 27 ++++++++++++++++++++++++++- 3 files changed, 34 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/block/genhd.c b/block/genhd.c index fda9c7a63c29..129ad939f9dd 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -653,7 +653,7 @@ void genhd_media_change_notify(struct gendisk *disk) EXPORT_SYMBOL_GPL(genhd_media_change_notify); #endif /* 0 */ -dev_t blk_lookup_devt(const char *name) +dev_t blk_lookup_devt(const char *name, int part) { struct device *dev; dev_t devt = MKDEV(0, 0); @@ -661,7 +661,11 @@ dev_t blk_lookup_devt(const char *name) mutex_lock(&block_class_lock); list_for_each_entry(dev, &block_class.devices, node) { if (strcmp(dev->bus_id, name) == 0) { - devt = dev->devt; + struct gendisk *disk = dev_to_disk(dev); + + if (part < disk->minors) + devt = MKDEV(MAJOR(dev->devt), + MINOR(dev->devt) + part); break; } } @@ -669,7 +673,6 @@ dev_t blk_lookup_devt(const char *name) return devt; } - EXPORT_SYMBOL(blk_lookup_devt); struct gendisk *alloc_disk(int minors) diff --git a/include/linux/genhd.h b/include/linux/genhd.h index e9874e7fcdf9..ae7aec3cabee 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -525,7 +525,7 @@ struct unixware_disklabel { #define ADDPART_FLAG_RAID 1 #define ADDPART_FLAG_WHOLEDISK 2 -extern dev_t blk_lookup_devt(const char *name); +extern dev_t blk_lookup_devt(const char *name, int part); extern char *disk_name (struct gendisk *hd, int part, char *buf); extern int rescan_partitions(struct gendisk *disk, struct block_device *bdev); @@ -553,7 +553,7 @@ static inline struct block_device *bdget_disk(struct gendisk *disk, int index) static inline void printk_all_partitions(void) { } -static inline dev_t blk_lookup_devt(const char *name) +static inline dev_t blk_lookup_devt(const char *name, int part) { dev_t devt = MKDEV(0, 0); return devt; diff --git a/init/do_mounts.c b/init/do_mounts.c index 3885e70e7759..660c1e50c91b 100644 --- a/init/do_mounts.c +++ b/init/do_mounts.c @@ -76,6 +76,7 @@ dev_t name_to_dev_t(char *name) char s[32]; char *p; dev_t res = 0; + int part; if (strncmp(name, "/dev/", 5) != 0) { unsigned maj, min; @@ -106,7 +107,31 @@ dev_t name_to_dev_t(char *name) for (p = s; *p; p++) if (*p == '/') *p = '!'; - res = blk_lookup_devt(s); + res = blk_lookup_devt(s, 0); + if (res) + goto done; + + /* + * try non-existant, but valid partition, which may only exist + * after revalidating the disk, like partitioned md devices + */ + while (p > s && isdigit(p[-1])) + p--; + if (p == s || !*p || *p == '0') + goto fail; + + /* try disk name without */ + part = simple_strtoul(p, NULL, 10); + *p = '\0'; + res = blk_lookup_devt(s, part); + if (res) + goto done; + + /* try disk name without p */ + if (p < s + 2 || !isdigit(p[-2]) || p[-1] != 'p') + goto fail; + p[-1] = '\0'; + res = blk_lookup_devt(s, part); if (res) goto done; -- cgit From 0a3ad00ca09632c6d0675f606276e92bdf1b306c Mon Sep 17 00:00:00 2001 From: Dave Young Date: Fri, 9 May 2008 15:24:08 +0800 Subject: Driver core: struct class remove children list because of the class_device was removed, now do the children list removing Signed-off-by: Dave Young Signed-off-by: Greg Kroah-Hartman --- drivers/base/class.c | 1 - include/linux/device.h | 1 - 2 files changed, 2 deletions(-) (limited to 'include') diff --git a/drivers/base/class.c b/drivers/base/class.c index 0ef00e8d4153..e085af0ff94f 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -140,7 +140,6 @@ int class_register(struct class *cls) pr_debug("device class '%s': registering\n", cls->name); - INIT_LIST_HEAD(&cls->children); INIT_LIST_HEAD(&cls->devices); INIT_LIST_HEAD(&cls->interfaces); kset_init(&cls->class_dirs); diff --git a/include/linux/device.h b/include/linux/device.h index 8c23e3dfe3ac..15e9fa3ad3af 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -183,7 +183,6 @@ struct class { struct module *owner; struct kset subsys; - struct list_head children; struct list_head devices; struct list_head interfaces; struct kset class_dirs; -- cgit From e0b4eb5193fed5c63413b0c137be29b0477d15ca Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Wed, 14 May 2008 23:06:15 +0200 Subject: make ide-iops.c:SELECT_MASK() static SELECT_MASK() can now become static. [bart: remove space between function name and open parenthesis] Signed-off-by: Adrian Bunk Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/ide-iops.c | 2 +- include/linux/ide.h | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c index 57d9a9a79a6f..0daf923541ff 100644 --- a/drivers/ide/ide-iops.c +++ b/drivers/ide/ide-iops.c @@ -95,7 +95,7 @@ void SELECT_DRIVE (ide_drive_t *drive) hwif->OUTB(drive->select.all, hwif->io_ports.device_addr); } -void SELECT_MASK (ide_drive_t *drive, int mask) +static void SELECT_MASK(ide_drive_t *drive, int mask) { const struct ide_port_ops *port_ops = drive->hwif->port_ops; diff --git a/include/linux/ide.h b/include/linux/ide.h index b0135b0c3a04..19ec852dffd2 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -965,7 +965,6 @@ typedef struct ide_task_s { void ide_tf_dump(const char *, struct ide_taskfile *); extern void SELECT_DRIVE(ide_drive_t *); -extern void SELECT_MASK(ide_drive_t *, int); extern int drive_is_ready(ide_drive_t *); -- cgit From cafa027b8cc6f605ccebc43a960644307a12d8dd Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Wed, 14 May 2008 23:06:16 +0200 Subject: cs5520: disable VDMA Disable Virtual DMA support for now (it causes system hangs). Thanks to TAKADA Yoshihito for the help with debugging the problem. Reported-by: TAKADA Yoshihito Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/pci/cs5520.c | 2 +- include/linux/ide.h | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/drivers/ide/pci/cs5520.c b/drivers/ide/pci/cs5520.c index 17669a434438..992b1cf8db69 100644 --- a/drivers/ide/pci/cs5520.c +++ b/drivers/ide/pci/cs5520.c @@ -119,6 +119,7 @@ static const struct ide_dma_ops cs5520_dma_ops = { .dma_timeout = ide_dma_timeout, }; +/* FIXME: VDMA is disabled because it caused system hangs */ #define DECLARE_CS_DEV(name_str) \ { \ .name = name_str, \ @@ -126,7 +127,6 @@ static const struct ide_dma_ops cs5520_dma_ops = { .dma_ops = &cs5520_dma_ops, \ .host_flags = IDE_HFLAG_ISA_PORTS | \ IDE_HFLAG_CS5520 | \ - IDE_HFLAG_VDMA | \ IDE_HFLAG_NO_ATAPI_DMA | \ IDE_HFLAG_ABUSE_SET_DMA_MODE, \ .pio_mask = ATA_PIO4, \ diff --git a/include/linux/ide.h b/include/linux/ide.h index 19ec852dffd2..f8f195c20da2 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1057,8 +1057,8 @@ enum { IDE_HFLAG_NO_SET_MODE = (1 << 9), /* trust BIOS for programming chipset/device for DMA */ IDE_HFLAG_TRUST_BIOS_FOR_DMA = (1 << 10), - /* host uses VDMA (tied with IDE_HFLAG_CS5520 for now) */ - IDE_HFLAG_VDMA = (1 << 11), + /* host is CS5510/CS5520 */ + IDE_HFLAG_CS5520 = (1 << 11), /* ATAPI DMA is unsupported */ IDE_HFLAG_NO_ATAPI_DMA = (1 << 12), /* set if host is a "non-bootable" controller */ @@ -1069,8 +1069,6 @@ enum { IDE_HFLAG_NO_AUTODMA = (1 << 15), /* host uses MMIO */ IDE_HFLAG_MMIO = (1 << 16), - /* host is CS5510/CS5520 */ - IDE_HFLAG_CS5520 = IDE_HFLAG_VDMA, /* no LBA48 */ IDE_HFLAG_NO_LBA48 = (1 << 17), /* no LBA48 DMA */ @@ -1100,6 +1098,8 @@ enum { IDE_HFLAG_NO_IO_32BIT = (1 << 30), /* never unmask IRQs */ IDE_HFLAG_NO_UNMASK_IRQS = (1 << 31), + /* host uses VDMA (disabled for now) */ + IDE_HFLAG_VDMA = 0, }; #ifdef CONFIG_BLK_DEV_OFFBOARD -- cgit From 7868f1ed84696d3cca83558e8dd459f20a36d077 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Tue, 6 May 2008 15:18:55 -0500 Subject: [IA64] Add header files for SGI UV platform Add new UV-specific header files. Signed-off-by: Jack Steiner Signed-off-by: Tony Luck --- include/asm-ia64/uv/uv_hub.h | 309 ++++++++++++++++++++++++++++++++++++++++++ include/asm-ia64/uv/uv_mmrs.h | 266 ++++++++++++++++++++++++++++++++++++ 2 files changed, 575 insertions(+) create mode 100644 include/asm-ia64/uv/uv_hub.h create mode 100644 include/asm-ia64/uv/uv_mmrs.h (limited to 'include') diff --git a/include/asm-ia64/uv/uv_hub.h b/include/asm-ia64/uv/uv_hub.h new file mode 100644 index 000000000000..f607018af4a1 --- /dev/null +++ b/include/asm-ia64/uv/uv_hub.h @@ -0,0 +1,309 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV architectural definitions + * + * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef __ASM_IA64_UV_HUB_H__ +#define __ASM_IA64_UV_HUB_H__ + +#include +#include +#include +#include + + +/* + * Addressing Terminology + * + * M - The low M bits of a physical address represent the offset + * into the blade local memory. RAM memory on a blade is physically + * contiguous (although various IO spaces may punch holes in + * it).. + * + * N - Number of bits in the node portion of a socket physical + * address. + * + * NASID - network ID of a router, Mbrick or Cbrick. Nasid values of + * routers always have low bit of 1, C/MBricks have low bit + * equal to 0. Most addressing macros that target UV hub chips + * right shift the NASID by 1 to exclude the always-zero bit. + * NASIDs contain up to 15 bits. + * + * GNODE - NASID right shifted by 1 bit. Most mmrs contain gnodes instead + * of nasids. + * + * PNODE - the low N bits of the GNODE. The PNODE is the most useful variant + * of the nasid for socket usage. + * + * + * NumaLink Global Physical Address Format: + * +--------------------------------+---------------------+ + * |00..000| GNODE | NodeOffset | + * +--------------------------------+---------------------+ + * |<-------53 - M bits --->|<--------M bits -----> + * + * M - number of node offset bits (35 .. 40) + * + * + * Memory/UV-HUB Processor Socket Address Format: + * +----------------+---------------+---------------------+ + * |00..000000000000| PNODE | NodeOffset | + * +----------------+---------------+---------------------+ + * <--- N bits --->|<--------M bits -----> + * + * M - number of node offset bits (35 .. 40) + * N - number of PNODE bits (0 .. 10) + * + * Note: M + N cannot currently exceed 44 (x86_64) or 46 (IA64). + * The actual values are configuration dependent and are set at + * boot time. M & N values are set by the hardware/BIOS at boot. + */ + + +/* + * Maximum number of bricks in all partitions and in all coherency domains. + * This is the total number of bricks accessible in the numalink fabric. It + * includes all C & M bricks. Routers are NOT included. + * + * This value is also the value of the maximum number of non-router NASIDs + * in the numalink fabric. + * + * NOTE: a brick may contain 1 or 2 OS nodes. Don't get these confused. + */ +#define UV_MAX_NUMALINK_BLADES 16384 + +/* + * Maximum number of C/Mbricks within a software SSI (hardware may support + * more). + */ +#define UV_MAX_SSI_BLADES 1 + +/* + * The largest possible NASID of a C or M brick (+ 2) + */ +#define UV_MAX_NASID_VALUE (UV_MAX_NUMALINK_NODES * 2) + +/* + * The following defines attributes of the HUB chip. These attributes are + * frequently referenced and are kept in the per-cpu data areas of each cpu. + * They are kept together in a struct to minimize cache misses. + */ +struct uv_hub_info_s { + unsigned long global_mmr_base; + unsigned long gpa_mask; + unsigned long gnode_upper; + unsigned long lowmem_remap_top; + unsigned long lowmem_remap_base; + unsigned short pnode; + unsigned short pnode_mask; + unsigned short coherency_domain_number; + unsigned short numa_blade_id; + unsigned char blade_processor_id; + unsigned char m_val; + unsigned char n_val; +}; +DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); +#define uv_hub_info (&__get_cpu_var(__uv_hub_info)) +#define uv_cpu_hub_info(cpu) (&per_cpu(__uv_hub_info, cpu)) + +/* + * Local & Global MMR space macros. + * Note: macros are intended to be used ONLY by inline functions + * in this file - not by other kernel code. + * n - NASID (full 15-bit global nasid) + * g - GNODE (full 15-bit global nasid, right shifted 1) + * p - PNODE (local part of nsids, right shifted 1) + */ +#define UV_NASID_TO_PNODE(n) (((n) >> 1) & uv_hub_info->pnode_mask) +#define UV_PNODE_TO_NASID(p) (((p) << 1) | uv_hub_info->gnode_upper) + +#define UV_LOCAL_MMR_BASE 0xf4000000UL +#define UV_GLOBAL_MMR32_BASE 0xf8000000UL +#define UV_GLOBAL_MMR64_BASE (uv_hub_info->global_mmr_base) + +#define UV_GLOBAL_MMR32_PNODE_SHIFT 15 +#define UV_GLOBAL_MMR64_PNODE_SHIFT 26 + +#define UV_GLOBAL_MMR32_PNODE_BITS(p) ((p) << (UV_GLOBAL_MMR32_PNODE_SHIFT)) + +#define UV_GLOBAL_MMR64_PNODE_BITS(p) \ + ((unsigned long)(p) << UV_GLOBAL_MMR64_PNODE_SHIFT) + +/* + * Macros for converting between kernel virtual addresses, socket local physical + * addresses, and UV global physical addresses. + * Note: use the standard __pa() & __va() macros for converting + * between socket virtual and socket physical addresses. + */ + +/* socket phys RAM --> UV global physical address */ +static inline unsigned long uv_soc_phys_ram_to_gpa(unsigned long paddr) +{ + if (paddr < uv_hub_info->lowmem_remap_top) + paddr += uv_hub_info->lowmem_remap_base; + return paddr | uv_hub_info->gnode_upper; +} + + +/* socket virtual --> UV global physical address */ +static inline unsigned long uv_gpa(void *v) +{ + return __pa(v) | uv_hub_info->gnode_upper; +} + +/* socket virtual --> UV global physical address */ +static inline void *uv_vgpa(void *v) +{ + return (void *)uv_gpa(v); +} + +/* UV global physical address --> socket virtual */ +static inline void *uv_va(unsigned long gpa) +{ + return __va(gpa & uv_hub_info->gpa_mask); +} + +/* pnode, offset --> socket virtual */ +static inline void *uv_pnode_offset_to_vaddr(int pnode, unsigned long offset) +{ + return __va(((unsigned long)pnode << uv_hub_info->m_val) | offset); +} + + +/* + * Access global MMRs using the low memory MMR32 space. This region supports + * faster MMR access but not all MMRs are accessible in this space. + */ +static inline unsigned long *uv_global_mmr32_address(int pnode, + unsigned long offset) +{ + return __va(UV_GLOBAL_MMR32_BASE | + UV_GLOBAL_MMR32_PNODE_BITS(pnode) | offset); +} + +static inline void uv_write_global_mmr32(int pnode, unsigned long offset, + unsigned long val) +{ + *uv_global_mmr32_address(pnode, offset) = val; +} + +static inline unsigned long uv_read_global_mmr32(int pnode, + unsigned long offset) +{ + return *uv_global_mmr32_address(pnode, offset); +} + +/* + * Access Global MMR space using the MMR space located at the top of physical + * memory. + */ +static inline unsigned long *uv_global_mmr64_address(int pnode, + unsigned long offset) +{ + return __va(UV_GLOBAL_MMR64_BASE | + UV_GLOBAL_MMR64_PNODE_BITS(pnode) | offset); +} + +static inline void uv_write_global_mmr64(int pnode, unsigned long offset, + unsigned long val) +{ + *uv_global_mmr64_address(pnode, offset) = val; +} + +static inline unsigned long uv_read_global_mmr64(int pnode, + unsigned long offset) +{ + return *uv_global_mmr64_address(pnode, offset); +} + +/* + * Access hub local MMRs. Faster than using global space but only local MMRs + * are accessible. + */ +static inline unsigned long *uv_local_mmr_address(unsigned long offset) +{ + return __va(UV_LOCAL_MMR_BASE | offset); +} + +static inline unsigned long uv_read_local_mmr(unsigned long offset) +{ + return *uv_local_mmr_address(offset); +} + +static inline void uv_write_local_mmr(unsigned long offset, unsigned long val) +{ + *uv_local_mmr_address(offset) = val; +} + +/* + * Structures and definitions for converting between cpu, node, pnode, and blade + * numbers. + */ + +/* Blade-local cpu number of current cpu. Numbered 0 .. <# cpus on the blade> */ +static inline int uv_blade_processor_id(void) +{ + return smp_processor_id(); +} + +/* Blade number of current cpu. Numnbered 0 .. <#blades -1> */ +static inline int uv_numa_blade_id(void) +{ + return 0; +} + +/* Convert a cpu number to the the UV blade number */ +static inline int uv_cpu_to_blade_id(int cpu) +{ + return 0; +} + +/* Convert linux node number to the UV blade number */ +static inline int uv_node_to_blade_id(int nid) +{ + return 0; +} + +/* Convert a blade id to the PNODE of the blade */ +static inline int uv_blade_to_pnode(int bid) +{ + return 0; +} + +/* Determine the number of possible cpus on a blade */ +static inline int uv_blade_nr_possible_cpus(int bid) +{ + return num_possible_cpus(); +} + +/* Determine the number of online cpus on a blade */ +static inline int uv_blade_nr_online_cpus(int bid) +{ + return num_online_cpus(); +} + +/* Convert a cpu id to the PNODE of the blade containing the cpu */ +static inline int uv_cpu_to_pnode(int cpu) +{ + return 0; +} + +/* Convert a linux node number to the PNODE of the blade */ +static inline int uv_node_to_pnode(int nid) +{ + return 0; +} + +/* Maximum possible number of blades */ +static inline int uv_num_possible_blades(void) +{ + return 1; +} + +#endif /* __ASM_IA64_UV_HUB__ */ + diff --git a/include/asm-ia64/uv/uv_mmrs.h b/include/asm-ia64/uv/uv_mmrs.h new file mode 100644 index 000000000000..1cc1dbb0182f --- /dev/null +++ b/include/asm-ia64/uv/uv_mmrs.h @@ -0,0 +1,266 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV MMR definitions + * + * Copyright (C) 2007-2008 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef __ASM_IA64_UV_MMRS__ +#define __ASM_IA64_UV_MMRS__ + +/* + * AUTO GENERATED - Do not edit + */ + + #define UV_MMR_ENABLE (1UL << 63) + +/* ========================================================================= */ +/* UVH_NODE_ID */ +/* ========================================================================= */ +#define UVH_NODE_ID 0x0UL + +#define UVH_NODE_ID_FORCE1_SHFT 0 +#define UVH_NODE_ID_FORCE1_MASK 0x0000000000000001UL +#define UVH_NODE_ID_MANUFACTURER_SHFT 1 +#define UVH_NODE_ID_MANUFACTURER_MASK 0x0000000000000ffeUL +#define UVH_NODE_ID_PART_NUMBER_SHFT 12 +#define UVH_NODE_ID_PART_NUMBER_MASK 0x000000000ffff000UL +#define UVH_NODE_ID_REVISION_SHFT 28 +#define UVH_NODE_ID_REVISION_MASK 0x00000000f0000000UL +#define UVH_NODE_ID_NODE_ID_SHFT 32 +#define UVH_NODE_ID_NODE_ID_MASK 0x00007fff00000000UL +#define UVH_NODE_ID_NODES_PER_BIT_SHFT 48 +#define UVH_NODE_ID_NODES_PER_BIT_MASK 0x007f000000000000UL +#define UVH_NODE_ID_NI_PORT_SHFT 56 +#define UVH_NODE_ID_NI_PORT_MASK 0x0f00000000000000UL + +union uvh_node_id_u { + unsigned long v; + struct uvh_node_id_s { + unsigned long force1 : 1; /* RO */ + unsigned long manufacturer : 11; /* RO */ + unsigned long part_number : 16; /* RO */ + unsigned long revision : 4; /* RO */ + unsigned long node_id : 15; /* RW */ + unsigned long rsvd_47 : 1; /* */ + unsigned long nodes_per_bit : 7; /* RW */ + unsigned long rsvd_55 : 1; /* */ + unsigned long ni_port : 4; /* RO */ + unsigned long rsvd_60_63 : 4; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR 0x16000d0UL + +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_SHFT 24 +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_MASK 0x00003fffff000000UL + +union uvh_rh_gam_alias210_redirect_config_0_mmr_u { + unsigned long v; + struct uvh_rh_gam_alias210_redirect_config_0_mmr_s { + unsigned long rsvd_0_23 : 24; /* */ + unsigned long dest_base : 22; /* RW */ + unsigned long rsvd_46_63: 18; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR 0x16000e0UL + +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR_DEST_BASE_SHFT 24 +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR_DEST_BASE_MASK 0x00003fffff000000UL + +union uvh_rh_gam_alias210_redirect_config_1_mmr_u { + unsigned long v; + struct uvh_rh_gam_alias210_redirect_config_1_mmr_s { + unsigned long rsvd_0_23 : 24; /* */ + unsigned long dest_base : 22; /* RW */ + unsigned long rsvd_46_63: 18; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR 0x16000f0UL + +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR_DEST_BASE_SHFT 24 +#define UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR_DEST_BASE_MASK 0x00003fffff000000UL + +union uvh_rh_gam_alias210_redirect_config_2_mmr_u { + unsigned long v; + struct uvh_rh_gam_alias210_redirect_config_2_mmr_s { + unsigned long rsvd_0_23 : 24; /* */ + unsigned long dest_base : 22; /* RW */ + unsigned long rsvd_46_63: 18; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR 0x1600010UL + +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_SHFT 28 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_MASK 0x00003ffff0000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_GR4_SHFT 46 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_GR4_MASK 0x0000400000000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_N_GRU_SHFT 52 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_N_GRU_MASK 0x00f0000000000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_ENABLE_SHFT 63 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_ENABLE_MASK 0x8000000000000000UL + +union uvh_rh_gam_gru_overlay_config_mmr_u { + unsigned long v; + struct uvh_rh_gam_gru_overlay_config_mmr_s { + unsigned long rsvd_0_27: 28; /* */ + unsigned long base : 18; /* RW */ + unsigned long gr4 : 1; /* RW */ + unsigned long rsvd_47_51: 5; /* */ + unsigned long n_gru : 4; /* RW */ + unsigned long rsvd_56_62: 7; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR 0x1600028UL + +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT 26 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_MASK 0x00003ffffc000000UL +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_DUAL_HUB_SHFT 46 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_DUAL_HUB_MASK 0x0000400000000000UL +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_ENABLE_SHFT 63 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_ENABLE_MASK 0x8000000000000000UL + +union uvh_rh_gam_mmr_overlay_config_mmr_u { + unsigned long v; + struct uvh_rh_gam_mmr_overlay_config_mmr_s { + unsigned long rsvd_0_25: 26; /* */ + unsigned long base : 20; /* RW */ + unsigned long dual_hub : 1; /* RW */ + unsigned long rsvd_47_62: 16; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RTC */ +/* ========================================================================= */ +#define UVH_RTC 0x28000UL + +#define UVH_RTC_REAL_TIME_CLOCK_SHFT 0 +#define UVH_RTC_REAL_TIME_CLOCK_MASK 0x00ffffffffffffffUL + +union uvh_rtc_u { + unsigned long v; + struct uvh_rtc_s { + unsigned long real_time_clock : 56; /* RW */ + unsigned long rsvd_56_63 : 8; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_SI_ADDR_MAP_CONFIG */ +/* ========================================================================= */ +#define UVH_SI_ADDR_MAP_CONFIG 0xc80000UL + +#define UVH_SI_ADDR_MAP_CONFIG_M_SKT_SHFT 0 +#define UVH_SI_ADDR_MAP_CONFIG_M_SKT_MASK 0x000000000000003fUL +#define UVH_SI_ADDR_MAP_CONFIG_N_SKT_SHFT 8 +#define UVH_SI_ADDR_MAP_CONFIG_N_SKT_MASK 0x0000000000000f00UL + +union uvh_si_addr_map_config_u { + unsigned long v; + struct uvh_si_addr_map_config_s { + unsigned long m_skt : 6; /* RW */ + unsigned long rsvd_6_7: 2; /* */ + unsigned long n_skt : 4; /* RW */ + unsigned long rsvd_12_63: 52; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_SI_ALIAS0_OVERLAY_CONFIG */ +/* ========================================================================= */ +#define UVH_SI_ALIAS0_OVERLAY_CONFIG 0xc80008UL + +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_BASE_SHFT 24 +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_BASE_MASK 0x00000000ff000000UL +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_M_ALIAS_SHFT 48 +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_M_ALIAS_MASK 0x001f000000000000UL +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_ENABLE_SHFT 63 +#define UVH_SI_ALIAS0_OVERLAY_CONFIG_ENABLE_MASK 0x8000000000000000UL + +union uvh_si_alias0_overlay_config_u { + unsigned long v; + struct uvh_si_alias0_overlay_config_s { + unsigned long rsvd_0_23: 24; /* */ + unsigned long base : 8; /* RW */ + unsigned long rsvd_32_47: 16; /* */ + unsigned long m_alias : 5; /* RW */ + unsigned long rsvd_53_62: 10; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_SI_ALIAS1_OVERLAY_CONFIG */ +/* ========================================================================= */ +#define UVH_SI_ALIAS1_OVERLAY_CONFIG 0xc80010UL + +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_BASE_SHFT 24 +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_BASE_MASK 0x00000000ff000000UL +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_M_ALIAS_SHFT 48 +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_M_ALIAS_MASK 0x001f000000000000UL +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_ENABLE_SHFT 63 +#define UVH_SI_ALIAS1_OVERLAY_CONFIG_ENABLE_MASK 0x8000000000000000UL + +union uvh_si_alias1_overlay_config_u { + unsigned long v; + struct uvh_si_alias1_overlay_config_s { + unsigned long rsvd_0_23: 24; /* */ + unsigned long base : 8; /* RW */ + unsigned long rsvd_32_47: 16; /* */ + unsigned long m_alias : 5; /* RW */ + unsigned long rsvd_53_62: 10; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_SI_ALIAS2_OVERLAY_CONFIG */ +/* ========================================================================= */ +#define UVH_SI_ALIAS2_OVERLAY_CONFIG 0xc80018UL + +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_BASE_SHFT 24 +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_BASE_MASK 0x00000000ff000000UL +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_M_ALIAS_SHFT 48 +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_M_ALIAS_MASK 0x001f000000000000UL +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_ENABLE_SHFT 63 +#define UVH_SI_ALIAS2_OVERLAY_CONFIG_ENABLE_MASK 0x8000000000000000UL + +union uvh_si_alias2_overlay_config_u { + unsigned long v; + struct uvh_si_alias2_overlay_config_s { + unsigned long rsvd_0_23: 24; /* */ + unsigned long base : 8; /* RW */ + unsigned long rsvd_32_47: 16; /* */ + unsigned long m_alias : 5; /* RW */ + unsigned long rsvd_53_62: 10; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + + +#endif /* __ASM_IA64_UV_MMRS__ */ -- cgit From 2224661494278bfc1c35b392cf6ee6f58e1d5e64 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Tue, 6 May 2008 15:18:57 -0500 Subject: [IA64] machvec support for SGI UV platform This patch adds the basic IA64 machvec infrastructure to support the SGI "UV" platform. Signed-off-by: Jack Steiner Signed-off-by: Tony Luck --- arch/ia64/Kconfig | 13 ++++++ arch/ia64/Makefile | 2 +- arch/ia64/kernel/acpi.c | 7 +++- arch/ia64/uv/Makefile | 12 ++++++ arch/ia64/uv/kernel/Makefile | 13 ++++++ arch/ia64/uv/kernel/machvec.c | 11 +++++ arch/ia64/uv/kernel/setup.c | 98 +++++++++++++++++++++++++++++++++++++++++++ include/asm-ia64/machvec.h | 2 + include/asm-ia64/machvec_uv.h | 26 ++++++++++++ 9 files changed, 182 insertions(+), 2 deletions(-) create mode 100644 arch/ia64/uv/Makefile create mode 100644 arch/ia64/uv/kernel/Makefile create mode 100644 arch/ia64/uv/kernel/machvec.c create mode 100644 arch/ia64/uv/kernel/setup.c create mode 100644 include/asm-ia64/machvec_uv.h (limited to 'include') diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 0df5f6f75edf..16be41446b5b 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -135,6 +135,7 @@ config IA64_GENERIC HP-zx1/sx1000 For HP systems HP-zx1/sx1000+swiotlb For HP systems with (broken) DMA-constrained devices. SGI-SN2 For SGI Altix systems + SGI-UV For SGI UV systems Ski-simulator For the HP simulator If you don't know what to do, choose "generic". @@ -170,6 +171,18 @@ config IA64_SGI_SN2 to select this option. If in doubt, select ia64 generic support instead. +config IA64_SGI_UV` + bool "SGI-UV`" + select NUMA + select ACPI_NUMA + select SWIOTLB + help + Selecting this option will optimize the kernel for use on UV based + systems, but the resulting kernel binary will not run on other + types of ia64 systems. If you have an SGI UV system, it's safe + to select this option. If in doubt, select ia64 generic support + instead. + config IA64_HP_SIM bool "Ski-simulator" select SWIOTLB diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile index ec4cca477f49..88f1a55c6c94 100644 --- a/arch/ia64/Makefile +++ b/arch/ia64/Makefile @@ -63,7 +63,7 @@ drivers-$(CONFIG_PCI) += arch/ia64/pci/ drivers-$(CONFIG_IA64_HP_SIM) += arch/ia64/hp/sim/ drivers-$(CONFIG_IA64_HP_ZX1) += arch/ia64/hp/common/ arch/ia64/hp/zx1/ drivers-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/hp/common/ arch/ia64/hp/zx1/ -drivers-$(CONFIG_IA64_GENERIC) += arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/ arch/ia64/sn/ +drivers-$(CONFIG_IA64_GENERIC) += arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/ arch/ia64/sn/ arch/ia64/uv/ drivers-$(CONFIG_OPROFILE) += arch/ia64/oprofile/ boot := arch/ia64/hp/sim/boot diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c index 19709a079635..6ff6815d34ef 100644 --- a/arch/ia64/kernel/acpi.c +++ b/arch/ia64/kernel/acpi.c @@ -117,7 +117,10 @@ acpi_get_sysname(void) if (!strcmp(hdr->oem_id, "HP")) { return "hpzx1"; } else if (!strcmp(hdr->oem_id, "SGI")) { - return "sn2"; + if (!strcmp(hdr->oem_table_id + 4, "UV")) + return "uv"; + else + return "sn2"; } return "dig"; @@ -130,6 +133,8 @@ acpi_get_sysname(void) return "hpzx1_swiotlb"; # elif defined (CONFIG_IA64_SGI_SN2) return "sn2"; +# elif defined (CONFIG_IA64_SGI_UV) + return "uv"; # elif defined (CONFIG_IA64_DIG) return "dig"; # else diff --git a/arch/ia64/uv/Makefile b/arch/ia64/uv/Makefile new file mode 100644 index 000000000000..aa9f91947c49 --- /dev/null +++ b/arch/ia64/uv/Makefile @@ -0,0 +1,12 @@ +# arch/ia64/uv/Makefile +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# +# Copyright (C) 2008 Silicon Graphics, Inc. All Rights Reserved. +# +# Makefile for the sn uv subplatform +# + +obj-y += kernel/ diff --git a/arch/ia64/uv/kernel/Makefile b/arch/ia64/uv/kernel/Makefile new file mode 100644 index 000000000000..8d92b4684d8e --- /dev/null +++ b/arch/ia64/uv/kernel/Makefile @@ -0,0 +1,13 @@ +# arch/ia64/uv/kernel/Makefile +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# +# Copyright (C) 2008 Silicon Graphics, Inc. All Rights Reserved. +# + +EXTRA_CFLAGS += -Iarch/ia64/sn/include + +obj-y += setup.o +obj-$(CONFIG_IA64_GENERIC) += machvec.o diff --git a/arch/ia64/uv/kernel/machvec.c b/arch/ia64/uv/kernel/machvec.c new file mode 100644 index 000000000000..50737a9dca74 --- /dev/null +++ b/arch/ia64/uv/kernel/machvec.c @@ -0,0 +1,11 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2008 Silicon Graphics, Inc. All Rights Reserved. + */ + +#define MACHVEC_PLATFORM_NAME uv +#define MACHVEC_PLATFORM_HEADER +#include diff --git a/arch/ia64/uv/kernel/setup.c b/arch/ia64/uv/kernel/setup.c new file mode 100644 index 000000000000..9aa743203c3c --- /dev/null +++ b/arch/ia64/uv/kernel/setup.c @@ -0,0 +1,98 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV Core Functions + * + * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved. + */ + +#include +#include +#include +#include +#include + +DEFINE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); +EXPORT_PER_CPU_SYMBOL_GPL(__uv_hub_info); + + +struct redir_addr { + unsigned long redirect; + unsigned long alias; +}; + +#define DEST_SHIFT UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_SHFT + +static __initdata struct redir_addr redir_addrs[] = { + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR, UVH_SI_ALIAS0_OVERLAY_CONFIG}, + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR, UVH_SI_ALIAS1_OVERLAY_CONFIG}, + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR, UVH_SI_ALIAS2_OVERLAY_CONFIG}, +}; + +static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size) +{ + union uvh_si_alias0_overlay_config_u alias; + union uvh_rh_gam_alias210_redirect_config_2_mmr_u redirect; + int i; + + for (i = 0; i < ARRAY_SIZE(redir_addrs); i++) { + alias.v = uv_read_local_mmr(redir_addrs[i].alias); + if (alias.s.base == 0) { + *size = (1UL << alias.s.m_alias); + redirect.v = uv_read_local_mmr(redir_addrs[i].redirect); + *base = (unsigned long)redirect.s.dest_base << DEST_SHIFT; + return; + } + } + BUG(); +} + +void __init uv_setup(char **cmdline_p) +{ + union uvh_si_addr_map_config_u m_n_config; + union uvh_node_id_u node_id; + unsigned long gnode_upper; + int nid, cpu, m_val, n_val; + unsigned long mmr_base, lowmem_redir_base, lowmem_redir_size; + + if (IS_MEDUSA()) { + lowmem_redir_base = 0; + lowmem_redir_size = 0; + node_id.v = 0; + m_n_config.s.m_skt = 37; + m_n_config.s.n_skt = 0; + mmr_base = 0; + } else { + get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); + node_id.v = uv_read_local_mmr(UVH_NODE_ID); + m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG); + mmr_base = + uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) & + ~UV_MMR_ENABLE; + } + + m_val = m_n_config.s.m_skt; + n_val = m_n_config.s.n_skt; + printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base); + + gnode_upper = (((unsigned long)node_id.s.node_id) & + ~((1 << n_val) - 1)) << m_val; + + for_each_present_cpu(cpu) { + nid = cpu_to_node(cpu); + uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base; + uv_cpu_hub_info(cpu)->lowmem_remap_top = + lowmem_redir_base + lowmem_redir_size; + uv_cpu_hub_info(cpu)->m_val = m_val; + uv_cpu_hub_info(cpu)->n_val = m_val; + uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) -1; + uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1; + uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper; + uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base; + uv_cpu_hub_info(cpu)->coherency_domain_number = 0;/* ZZZ */ + printk(KERN_DEBUG "UV cpu %d, nid %d\n", cpu, nid); + } +} + diff --git a/include/asm-ia64/machvec.h b/include/asm-ia64/machvec.h index 9f020eb825c5..0721a5e8271e 100644 --- a/include/asm-ia64/machvec.h +++ b/include/asm-ia64/machvec.h @@ -126,6 +126,8 @@ extern void machvec_tlb_migrate_finish (struct mm_struct *); # include # elif defined (CONFIG_IA64_SGI_SN2) # include +# elif defined (CONFIG_IA64_SGI_UV) +# include # elif defined (CONFIG_IA64_GENERIC) # ifdef MACHVEC_PLATFORM_HEADER diff --git a/include/asm-ia64/machvec_uv.h b/include/asm-ia64/machvec_uv.h new file mode 100644 index 000000000000..2931447f3813 --- /dev/null +++ b/include/asm-ia64/machvec_uv.h @@ -0,0 +1,26 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV Core Functions + * + * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef _ASM_IA64_MACHVEC_UV_H +#define _ASM_IA64_MACHVEC_UV_H + +extern ia64_mv_setup_t uv_setup; + +/* + * This stuff has dual use! + * + * For a generic kernel, the macros are used to initialize the + * platform's machvec structure. When compiling a non-generic kernel, + * the macros are used directly. + */ +#define platform_name "uv" +#define platform_setup uv_setup + +#endif /* _ASM_IA64_MACHVEC_UV_H */ -- cgit From b32a09db4fb9a87246ba4e7726a979ac4709ad97 Mon Sep 17 00:00:00 2001 From: Markus Armbruster Date: Tue, 26 Feb 2008 09:57:11 -0600 Subject: add match_strlcpy() us it to make v9fs make uname and remotename parsing more robust match_strcpy() is a somewhat creepy function: the caller needs to make sure that the destination buffer is big enough, and when he screws up or forgets, match_strcpy() happily overruns the buffer. There's exactly one customer: v9fs_parse_options(). I believe it currently can't overflow its buffer, but that's not exactly obvious. The source string is a substing of the mount options. The kernel silently truncates those to PAGE_SIZE bytes, including the terminating zero. See compat_sys_mount() and do_mount(). The destination buffer is obtained from __getname(), which allocates from name_cachep, which is initialized by vfs_caches_init() for size PATH_MAX. We're safe as long as PATH_MAX <= PAGE_SIZE. PATH_MAX is 4096. As far as I know, the smallest PAGE_SIZE is also 4096. Here's a patch that makes the code a bit more obviously correct. It doesn't depend on PATH_MAX <= PAGE_SIZE. Signed-off-by: Markus Armbruster Cc: Latchesar Ionkov Cc: Jim Meyering Cc: "Randy.Dunlap" Signed-off-by: Andrew Morton Signed-off-by: Eric Van Hensbergen --- fs/9p/v9fs.c | 4 ++-- include/linux/parser.h | 2 +- lib/parser.c | 32 ++++++++++++++++++++------------ 3 files changed, 23 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 9b0f0222e8bb..e307fbd34fa0 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -125,10 +125,10 @@ static void v9fs_parse_options(struct v9fs_session_info *v9ses) v9ses->afid = option; break; case Opt_uname: - match_strcpy(v9ses->uname, &args[0]); + match_strlcpy(v9ses->uname, &args[0], PATH_MAX); break; case Opt_remotename: - match_strcpy(v9ses->aname, &args[0]); + match_strlcpy(v9ses->aname, &args[0], PATH_MAX); break; case Opt_nodevmap: v9ses->nodev = 1; diff --git a/include/linux/parser.h b/include/linux/parser.h index 26b2bdfcaf06..7dcd05075756 100644 --- a/include/linux/parser.h +++ b/include/linux/parser.h @@ -29,5 +29,5 @@ int match_token(char *, match_table_t table, substring_t args[]); int match_int(substring_t *, int *result); int match_octal(substring_t *, int *result); int match_hex(substring_t *, int *result); -void match_strcpy(char *, const substring_t *); +size_t match_strlcpy(char *, const substring_t *, size_t); char *match_strdup(const substring_t *); diff --git a/lib/parser.c b/lib/parser.c index 703c8c13b346..4f0cbc03e0e8 100644 --- a/lib/parser.c +++ b/lib/parser.c @@ -182,18 +182,25 @@ int match_hex(substring_t *s, int *result) } /** - * match_strcpy: - copies the characters from a substring_t to a string - * @to: string to copy characters to. - * @s: &substring_t to copy + * match_strlcpy: - Copy the characters from a substring_t to a sized buffer + * @dest: where to copy to + * @src: &substring_t to copy + * @size: size of destination buffer * - * Description: Copies the set of characters represented by the given - * &substring_t @s to the c-style string @to. Caller guarantees that @to is - * large enough to hold the characters of @s. + * Description: Copy the characters in &substring_t @src to the + * c-style string @dest. Copy no more than @size - 1 characters, plus + * the terminating NUL. Return length of @src. */ -void match_strcpy(char *to, const substring_t *s) +size_t match_strlcpy(char *dest, const substring_t *src, size_t size) { - memcpy(to, s->from, s->to - s->from); - to[s->to - s->from] = '\0'; + size_t ret = src->to - src->from; + + if (size) { + size_t len = ret >= size ? size - 1 : ret; + memcpy(dest, src->from, len); + dest[len] = '\0'; + } + return ret; } /** @@ -206,9 +213,10 @@ void match_strcpy(char *to, const substring_t *s) */ char *match_strdup(const substring_t *s) { - char *p = kmalloc(s->to - s->from + 1, GFP_KERNEL); + size_t sz = s->to - s->from + 1; + char *p = kmalloc(sz, GFP_KERNEL); if (p) - match_strcpy(p, s); + match_strlcpy(p, s, sz); return p; } @@ -216,5 +224,5 @@ EXPORT_SYMBOL(match_token); EXPORT_SYMBOL(match_int); EXPORT_SYMBOL(match_octal); EXPORT_SYMBOL(match_hex); -EXPORT_SYMBOL(match_strcpy); +EXPORT_SYMBOL(match_strlcpy); EXPORT_SYMBOL(match_strdup); -- cgit From ee443996a35c1e04f210cafd43d5a98d41e46085 Mon Sep 17 00:00:00 2001 From: Eric Van Hensbergen Date: Wed, 5 Mar 2008 07:08:09 -0600 Subject: 9p: Documentation updates The kernel-doc comments of much of the 9p system have been in disarray since reorganization. This patch fixes those problems, adds additional documentation and a template book which collects the 9p information. Signed-off-by: Eric Van Hensbergen --- fs/9p/fid.h | 15 +++ fs/9p/v9fs.c | 8 +- fs/9p/v9fs.h | 85 +++++++++++----- fs/9p/vfs_addr.c | 2 +- fs/9p/vfs_dir.c | 2 +- fs/9p/vfs_file.c | 11 ++- fs/9p/vfs_inode.c | 50 +++++++--- fs/9p/vfs_super.c | 1 + include/net/9p/9p.h | 239 +++++++++++++++++++++++++++++++++++++++------ include/net/9p/client.h | 35 +++++++ include/net/9p/transport.h | 43 ++++++++ net/9p/conv.c | 128 +++++++++++++++++++++++- net/9p/error.c | 11 ++- net/9p/fcprint.c | 8 ++ net/9p/mod.c | 7 +- net/9p/trans_fd.c | 146 ++++++++++++++++++++++----- net/9p/trans_virtio.c | 155 +++++++++++++++++++++++++++-- net/9p/util.c | 32 ++++-- 18 files changed, 854 insertions(+), 124 deletions(-) (limited to 'include') diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 26e07df783b9..c3bbd6af996d 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -22,6 +22,21 @@ #include +/** + * struct v9fs_dentry - 9p private data stored in dentry d_fsdata + * @lock: protects the fidlist + * @fidlist: list of FIDs currently associated with this dentry + * + * This structure defines the 9p private data associated with + * a particular dentry. In particular, this private data is used + * to lookup which 9P FID handle should be used for a particular VFS + * operation. FID handles are associated with dentries instead of + * inodes in order to more closely map functionality to the Plan 9 + * expected behavior for FID reclaimation and tracking. + * + * See Also: Mapping FIDs to Linux VFS model in + * Design and Implementation of the Linux 9P File System documentation + */ struct v9fs_dentry { spinlock_t lock; /* protect fidlist */ struct list_head fidlist; diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index e307fbd34fa0..79d310c00188 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -71,7 +71,6 @@ static match_table_t tokens = { /** * v9fs_parse_options - parse mount options into session structure - * @options: options string passed from mount * @v9ses: existing v9fs session information * */ @@ -256,9 +255,12 @@ void v9fs_session_close(struct v9fs_session_info *v9ses) } /** - * v9fs_session_cancel - mark transport as disconnected - * and cancel all pending requests. + * v9fs_session_cancel - terminate a session + * @v9ses: session to terminate + * + * mark transport as disconnected and cancel all pending requests. */ + void v9fs_session_cancel(struct v9fs_session_info *v9ses) { P9_DPRINTK(P9_DEBUG_ERROR, "cancel session %p\n", v9ses); p9_client_disconnect(v9ses->clnt); diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index 7d3a1018db52..a7d567192998 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -21,18 +21,69 @@ * */ -/* - * Session structure provides information for an opened session - * - */ +/** + * enum p9_session_flags - option flags for each 9P session + * @V9FS_EXTENDED: whether or not to use 9P2000.u extensions + * @V9FS_ACCESS_SINGLE: only the mounting user can access the hierarchy + * @V9FS_ACCESS_USER: a new attach will be issued for every user (default) + * @V9FS_ACCESS_ANY: use a single attach for all users + * @V9FS_ACCESS_MASK: bit mask of different ACCESS options + * + * Session flags reflect options selected by users at mount time + */ +enum p9_session_flags { + V9FS_EXTENDED = 0x01, + V9FS_ACCESS_SINGLE = 0x02, + V9FS_ACCESS_USER = 0x04, + V9FS_ACCESS_ANY = 0x06, + V9FS_ACCESS_MASK = 0x06, +}; + +/* possible values of ->cache */ +/** + * enum p9_cache_modes - user specified cache preferences + * @CACHE_NONE: do not cache data, dentries, or directory contents (default) + * @CACHE_LOOSE: cache data, dentries, and directory contents w/no consistency + * + * eventually support loose, tight, time, session, default always none + */ + +enum p9_cache_modes { + CACHE_NONE, + CACHE_LOOSE, +}; + +/** + * struct v9fs_session_info - per-instance session information + * @flags: session options of type &p9_session_flags + * @nodev: set to 1 to disable device mapping + * @debug: debug level + * @afid: authentication handle + * @cache: cache mode of type &p9_cache_modes + * @options: copy of options string given by user + * @uname: string user name to mount hierarchy as + * @aname: mount specifier for remote hierarchy + * @maxdata: maximum data to be sent/recvd per protocol message + * @dfltuid: default numeric userid to mount hierarchy as + * @dfltgid: default numeric groupid to mount hierarchy as + * @uid: if %V9FS_ACCESS_SINGLE, the numeric uid which mounted the hierarchy + * @clnt: reference to 9P network client instantiated for this session + * @debugfs_dir: reference to debugfs_dir which can be used for add'l debug + * + * This structure holds state for each session instance established during + * a sys_mount() . + * + * Bugs: there seems to be a lot of state which could be condensed and/or + * removed. + */ struct v9fs_session_info { /* options */ - unsigned char flags; /* session flags */ - unsigned char nodev; /* set to 1 if no disable device mapping */ - unsigned short debug; /* debug level */ - unsigned int afid; /* authentication fid */ - unsigned int cache; /* cache mode */ + unsigned char flags; + unsigned char nodev; + unsigned short debug; + unsigned int afid; + unsigned int cache; char *options; /* copy of mount options */ char *uname; /* user name to mount as */ @@ -45,22 +96,6 @@ struct v9fs_session_info { struct dentry *debugfs_dir; }; -/* session flags */ -enum { - V9FS_EXTENDED = 0x01, /* 9P2000.u */ - V9FS_ACCESS_MASK = 0x06, /* access mask */ - V9FS_ACCESS_SINGLE = 0x02, /* only one user can access the files */ - V9FS_ACCESS_USER = 0x04, /* attache per user */ - V9FS_ACCESS_ANY = 0x06, /* use the same attach for all users */ -}; - -/* possible values of ->cache */ -/* eventually support loose, tight, time, session, default always none */ -enum { - CACHE_NONE, /* default */ - CACHE_LOOSE, /* no consistency */ -}; - extern struct dentry *v9fs_debugfs_root; struct p9_fid *v9fs_session_init(struct v9fs_session_info *, const char *, diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index 6248f0e727a3..97d3aed57983 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -43,7 +43,7 @@ /** * v9fs_vfs_readpage - read an entire page in from 9P * - * @file: file being read + * @filp: file being read * @page: structure to page * */ diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index 0924d4477da3..88e3787c6ea9 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -60,7 +60,7 @@ static inline int dt_type(struct p9_stat *mistat) /** * v9fs_dir_readdir - read a directory - * @filep: opened file structure + * @filp: opened file structure * @dirent: directory structure ??? * @filldir: function to populate directory structure ??? * diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index a616fff8906d..0d55affe37d4 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -90,10 +90,11 @@ int v9fs_file_open(struct inode *inode, struct file *file) /** * v9fs_file_lock - lock a file (or directory) - * @inode: inode to be opened - * @file: file being opened + * @filp: file to be locked + * @cmd: lock command + * @fl: file lock structure * - * XXX - this looks like a local only lock, we should extend into 9P + * Bugs: this looks like a local only lock, we should extend into 9P * by using open exclusive */ @@ -118,7 +119,7 @@ static int v9fs_file_lock(struct file *filp, int cmd, struct file_lock *fl) /** * v9fs_file_read - read from a file - * @filep: file pointer to read + * @filp: file pointer to read * @data: data buffer to read data into * @count: size of buffer * @offset: offset at which to read data @@ -142,7 +143,7 @@ v9fs_file_read(struct file *filp, char __user * data, size_t count, /** * v9fs_file_write - write to a file - * @filep: file pointer to write + * @filp: file pointer to write * @data: data buffer to write data from * @count: size of buffer * @offset: offset at which to write data diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 6a28842052ea..40fa807bd929 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -129,6 +129,12 @@ static int p9mode2unixmode(struct v9fs_session_info *v9ses, int mode) return res; } +/** + * v9fs_uflags2omode- convert posix open flags to plan 9 mode bits + * @uflags: flags to convert + * + */ + int v9fs_uflags2omode(int uflags) { int ret; @@ -312,6 +318,14 @@ error: } */ +/** + * v9fs_inode_from_fid - populate an inode by issuing a attribute request + * @v9ses: session information + * @fid: fid to issue attribute request for + * @sb: superblock on which to create inode + * + */ + static struct inode * v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, struct super_block *sb) @@ -384,9 +398,12 @@ v9fs_open_created(struct inode *inode, struct file *file) /** * v9fs_create - Create a file + * @v9ses: session information + * @dir: directory that dentry is being created in * @dentry: dentry that is being created * @perm: create permissions * @mode: open mode + * @extension: 9p2000.u extension string to support devices, etc. * */ static struct p9_fid * @@ -461,7 +478,7 @@ error: /** * v9fs_vfs_create - VFS hook to create files - * @inode: directory inode that is being created + * @dir: directory inode that is being created * @dentry: dentry that is being deleted * @mode: create permissions * @nd: path information @@ -519,7 +536,7 @@ error: /** * v9fs_vfs_mkdir - VFS mkdir hook to create a directory - * @inode: inode that is being unlinked + * @dir: inode that is being unlinked * @dentry: dentry that is being unlinked * @mode: mode for new directory * @@ -703,9 +720,9 @@ done: /** * v9fs_vfs_getattr - retrieve file metadata - * @mnt - mount information - * @dentry - file to get attributes on - * @stat - metadata structure to populate + * @mnt: mount information + * @dentry: file to get attributes on + * @stat: metadata structure to populate * */ @@ -928,7 +945,7 @@ done: /** * v9fs_vfs_readlink - read a symlink's location * @dentry: dentry for symlink - * @buf: buffer to load symlink location into + * @buffer: buffer to load symlink location into * @buflen: length of buffer * */ @@ -996,10 +1013,12 @@ static void *v9fs_vfs_follow_link(struct dentry *dentry, struct nameidata *nd) * v9fs_vfs_put_link - release a symlink path * @dentry: dentry for symlink * @nd: nameidata + * @p: unused * */ -static void v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd, void *p) +static void +v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd, void *p) { char *s = nd_get_link(nd); @@ -1008,6 +1027,15 @@ static void v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd, void __putname(s); } +/** + * v9fs_vfs_mkspecial - create a special file + * @dir: inode to create special file in + * @dentry: dentry to create + * @mode: mode to create special file + * @extension: 9p2000.u format extension string representing special file + * + */ + static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, int mode, const char *extension) { @@ -1037,7 +1065,7 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, * @dentry: dentry for symlink * @symname: symlink data * - * See 9P2000.u RFC for more information + * See Also: 9P2000.u RFC for more information * */ @@ -1058,10 +1086,6 @@ v9fs_vfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) * */ -/* XXX - lots of code dup'd from symlink and creates, - * figure out a better reuse strategy - */ - static int v9fs_vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry) @@ -1098,7 +1122,7 @@ clunk_fid: * @dir: inode destination for new link * @dentry: dentry for file * @mode: mode for creation - * @dev_t: device associated with special file + * @rdev: device associated with special file * */ diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index a452ac67fc94..ba10f172626d 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -75,6 +75,7 @@ static int v9fs_set_super(struct super_block *s, void *data) * v9fs_fill_super - populate superblock with info * @sb: superblock * @v9ses: session information + * @flags: flags propagated from v9fs_get_sb() * */ diff --git a/include/net/9p/9p.h b/include/net/9p/9p.h index 585eb4496990..7bfb2f2e423c 100644 --- a/include/net/9p/9p.h +++ b/include/net/9p/9p.h @@ -29,14 +29,31 @@ #ifdef CONFIG_NET_9P_DEBUG -#define P9_DEBUG_ERROR (1<<0) -#define P9_DEBUG_9P (1<<2) -#define P9_DEBUG_VFS (1<<3) -#define P9_DEBUG_CONV (1<<4) -#define P9_DEBUG_MUX (1<<5) -#define P9_DEBUG_TRANS (1<<6) -#define P9_DEBUG_SLABS (1<<7) -#define P9_DEBUG_FCALL (1<<8) +/** + * enum p9_debug_flags - bits for mount time debug parameter + * @P9_DEBUG_ERROR: more verbose error messages including original error string + * @P9_DEBUG_9P: 9P protocol tracing + * @P9_DEBUG_VFS: VFS API tracing + * @P9_DEBUG_CONV: protocol conversion tracing + * @P9_DEBUG_MUX: trace management of concurrent transactions + * @P9_DEBUG_TRANS: transport tracing + * @P9_DEBUG_SLABS: memory management tracing + * @P9_DEBUG_FCALL: verbose dump of protocol messages + * + * These flags are passed at mount time to turn on various levels of + * verbosity and tracing which will be output to the system logs. + */ + +enum p9_debug_flags { + P9_DEBUG_ERROR = (1<<0), + P9_DEBUG_9P = (1<<2), + P9_DEBUG_VFS = (1<<3), + P9_DEBUG_CONV = (1<<4), + P9_DEBUG_MUX = (1<<5), + P9_DEBUG_TRANS = (1<<6), + P9_DEBUG_SLABS = (1<<7), + P9_DEBUG_FCALL = (1<<8), +}; extern unsigned int p9_debug_level; @@ -62,9 +79,47 @@ do { \ format , __FUNCTION__, task_pid_nr(current), ## arg); \ } while (0) +/** + * enum p9_msg_t - 9P message types + * @P9_TVERSION: version handshake request + * @P9_RVERSION: version handshake response + * @P9_TAUTH: request to establish authentication channel + * @P9_RAUTH: response with authentication information + * @P9_TATTACH: establish user access to file service + * @P9_RATTACH: response with top level handle to file hierarchy + * @P9_TERROR: not used + * @P9_RERROR: response for any failed request + * @P9_TFLUSH: request to abort a previous request + * @P9_RFLUSH: response when previous request has been cancelled + * @P9_TWALK: descend a directory hierarchy + * @P9_RWALK: response with new handle for position within hierarchy + * @P9_TOPEN: prepare a handle for I/O on an existing file + * @P9_ROPEN: response with file access information + * @P9_TCREATE: prepare a handle for I/O on a new file + * @P9_RCREATE: response with file access information + * @P9_TREAD: request to transfer data from a file or directory + * @P9_RREAD: response with data requested + * @P9_TWRITE: reuqest to transfer data to a file + * @P9_RWRITE: response with out much data was transfered to file + * @P9_TCLUNK: forget about a handle to an entity within the file system + * @P9_RCLUNK: response when server has forgotten about the handle + * @P9_TREMOVE: request to remove an entity from the hierarchy + * @P9_RREMOVE: response when server has removed the entity + * @P9_TSTAT: request file entity attributes + * @P9_RSTAT: response with file entity attributes + * @P9_TWSTAT: request to update file entity attributes + * @P9_RWSTAT: response when file entity attributes are updated + * + * There are 14 basic operations in 9P2000, paired as + * requests and responses. The one special case is ERROR + * as there is no @P9_TERROR request for clients to transmit to + * the server, but the server may respond to any other request + * with an @P9_RERROR. + * + * See Also: http://plan9.bell-labs.com/sys/man/5/INDEX.html + */ -/* Message Types */ -enum { +enum p9_msg_t { P9_TVERSION = 100, P9_RVERSION, P9_TAUTH = 102, @@ -95,30 +150,71 @@ enum { P9_RWSTAT, }; -/* open modes */ -enum { +/** + * enum p9_open_mode_t - 9P open modes + * @P9_OREAD: open file for reading only + * @P9_OWRITE: open file for writing only + * @P9_ORDWR: open file for reading or writing + * @P9_OEXEC: open file for execution + * @P9_OTRUNC: truncate file to zero-length before opening it + * @P9_OREXEC: close the file when an exec(2) system call is made + * @P9_ORCLOSE: remove the file when the file is closed + * @P9_OAPPEND: open the file and seek to the end + * @P9_OEXCL: only create a file, do not open it + * + * 9P open modes differ slightly from Posix standard modes. + * In particular, there are extra modes which specify different + * semantic behaviors than may be available on standard Posix + * systems. For example, @P9_OREXEC and @P9_ORCLOSE are modes that + * most likely will not be issued from the Linux VFS client, but may + * be supported by servers. + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/open + */ + +enum p9_open_mode_t { P9_OREAD = 0x00, P9_OWRITE = 0x01, P9_ORDWR = 0x02, P9_OEXEC = 0x03, - P9_OEXCL = 0x04, P9_OTRUNC = 0x10, P9_OREXEC = 0x20, P9_ORCLOSE = 0x40, P9_OAPPEND = 0x80, -}; - -/* permissions */ -enum { + P9_OEXCL = 0x1000, +}; + +/** + * enum p9_perm_t - 9P permissions + * @P9_DMDIR: mode bite for directories + * @P9_DMAPPEND: mode bit for is append-only + * @P9_DMEXCL: mode bit for excluse use (only one open handle allowed) + * @P9_DMMOUNT: mode bite for mount points + * @P9_DMAUTH: mode bit for authentication file + * @P9_DMTMP: mode bit for non-backed-up files + * @P9_DMSYMLINK: mode bit for symbolic links (9P2000.u) + * @P9_DMLINK: mode bit for hard-link (9P2000.u) + * @P9_DMDEVICE: mode bit for device files (9P2000.u) + * @P9_DMNAMEDPIPE: mode bit for named pipe (9P2000.u) + * @P9_DMSOCKET: mode bit for socket (9P2000.u) + * @P9_DMSETUID: mode bit for setuid (9P2000.u) + * @P9_DMSETGID: mode bit for setgid (9P2000.u) + * @P9_DMSETVTX: mode bit for sticky bit (9P2000.u) + * + * 9P permissions differ slightly from Posix standard modes. + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/stat + */ +enum p9_perm_t { P9_DMDIR = 0x80000000, P9_DMAPPEND = 0x40000000, P9_DMEXCL = 0x20000000, P9_DMMOUNT = 0x10000000, P9_DMAUTH = 0x08000000, P9_DMTMP = 0x04000000, +/* 9P2000.u extensions */ P9_DMSYMLINK = 0x02000000, P9_DMLINK = 0x01000000, - /* 9P2000.u extensions */ P9_DMDEVICE = 0x00800000, P9_DMNAMEDPIPE = 0x00200000, P9_DMSOCKET = 0x00100000, @@ -127,8 +223,26 @@ enum { P9_DMSETVTX = 0x00010000, }; -/* qid.types */ -enum { +/** + * enum p9_qid_t - QID types + * @P9_QTDIR: directory + * @P9_QTAPPEND: append-only + * @P9_QTEXCL: excluse use (only one open handle allowed) + * @P9_QTMOUNT: mount points + * @P9_QTAUTH: authentication file + * @P9_QTTMP: non-backed-up files + * @P9_QTSYMLINK: symbolic links (9P2000.u) + * @P9_QTLINK: hard-link (9P2000.u) + * @P9_QTFILE: normal files + * + * QID types are a subset of permissions - they are primarily + * used to differentiate semantics for a file system entity via + * a jump-table. Their value is also the most signifigant 16 bits + * of the permission_t + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/stat + */ +enum p9_qid_t { P9_QTDIR = 0x80, P9_QTAPPEND = 0x40, P9_QTEXCL = 0x20, @@ -140,6 +254,7 @@ enum { P9_QTFILE = 0x00, }; +/* 9P Magic Numbers */ #define P9_NOTAG (u16)(~0) #define P9_NOFID (u32)(~0) #define P9_MAXWELEM 16 @@ -147,19 +262,69 @@ enum { /* ample room for Twrite/Rread header */ #define P9_IOHDRSZ 24 +/** + * struct p9_str - length prefixed string type + * @len: length of the string + * @str: the string + * + * The protocol uses length prefixed strings for all + * string data, so we replicate that for our internal + * string members. + */ + struct p9_str { u16 len; char *str; }; -/* qids are the unique ID for a file (like an inode */ +/** + * struct p9_qid - file system entity information + * @type: 8-bit type &p9_qid_t + * @version: 16-bit monotonically incrementing version number + * @path: 64-bit per-server-unique ID for a file system element + * + * qids are identifiers used by 9P servers to track file system + * entities. The type is used to differentiate semantics for operations + * on the entity (ie. read means something different on a directory than + * on a file). The path provides a server unique index for an entity + * (roughly analogous to an inode number), while the version is updated + * every time a file is modified and can be used to maintain cache + * coherency between clients and serves. + * Servers will often differentiate purely synthetic entities by setting + * their version to 0, signaling that they should never be cached and + * should be accessed synchronously. + * + * See Also://plan9.bell-labs.com/magic/man2html/2/stat + */ + struct p9_qid { u8 type; u32 version; u64 path; }; -/* Plan 9 file metadata (stat) structure */ +/** + * struct p9_stat - file system metadata information + * @size: length prefix for this stat structure instance + * @type: the type of the server (equivilent to a major number) + * @dev: the sub-type of the server (equivilent to a minor number) + * @qid: unique id from the server of type &p9_qid + * @mode: Plan 9 format permissions of type &p9_perm_t + * @atime: Last access/read time + * @mtime: Last modify/write time + * @length: file length + * @name: last element of path (aka filename) in type &p9_str + * @uid: owner name in type &p9_str + * @gid: group owner in type &p9_str + * @muid: last modifier in type &p9_str + * @extension: area used to encode extended UNIX support in type &p9_str + * @n_uid: numeric user id of owner (part of 9p2000.u extension) + * @n_gid: numeric group id (part of 9p2000.u extension) + * @n_muid: numeric user id of laster modifier (part of 9p2000.u extension) + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/stat + */ + struct p9_stat { u16 size; u16 type; @@ -179,10 +344,14 @@ struct p9_stat { u32 n_muid; /* 9p2000.u extensions */ }; -/* file metadata (stat) structure used to create Twstat message - The is similar to p9_stat, but the strings don't point to - the same memory block and should be freed separately -*/ +/* + * file metadata (stat) structure used to create Twstat message + * The is identical to &p9_stat, but the strings don't point to + * the same memory block and should be freed separately + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/stat + */ + struct p9_wstat { u16 size; u16 type; @@ -335,10 +504,20 @@ struct p9_twstat { struct p9_rwstat { }; -/* - * fcall is the primary packet structure - * - */ +/** + * struct p9_fcall - primary packet structure + * @size: prefixed length of the structure + * @id: protocol operating identifier of type &p9_msg_t + * @tag: transaction id of the request + * @sdata: payload + * @params: per-operation parameters + * + * &p9_fcall represents the structure for all 9P RPC + * transactions. Requests are packaged into fcalls, and reponses + * must be extracted from them. + * + * See Also: http://plan9.bell-labs.com/magic/man2html/2/fcall + */ struct p9_fcall { u32 size; diff --git a/include/net/9p/client.h b/include/net/9p/client.h index e52f93d9ac5f..c936dd14de41 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -26,6 +26,23 @@ #ifndef NET_9P_CLIENT_H #define NET_9P_CLIENT_H +/** + * struct p9_client - per client instance state + * @lock: protect @fidlist + * @msize: maximum data size negotiated by protocol + * @dotu: extension flags negotiated by protocol + * @trans_mod: module API instantiated with this client + * @trans: tranport instance state and API + * @conn: connection state information used by trans_fd + * @fidpool: fid handle accounting for session + * @fidlist: List of active fid handles + * + * The client structure is used to keep track of various per-client + * state that has been instantiated. + * + * Bugs: duplicated data and potentially unnecessary elements. + */ + struct p9_client { spinlock_t lock; /* protect client structure */ int msize; @@ -38,6 +55,24 @@ struct p9_client { struct list_head fidlist; }; +/** + * struct p9_fid - file system entity handle + * @clnt: back pointer to instantiating &p9_client + * @fid: numeric identifier for this handle + * @mode: current mode of this fid (enum?) + * @qid: the &p9_qid server identifier this handle points to + * @iounit: the server reported maximum transaction size for this file + * @uid: the numeric uid of the local user who owns this handle + * @aux: transport specific information (unused?) + * @rdir_fpos: tracks offset of file position when reading directory contents + * @rdir_pos: (unused?) + * @rdir_fcall: holds response of last directory read request + * @flist: per-client-instance fid tracking + * @dlist: per-dentry fid tracking + * + * TODO: This needs lots of explanation. + */ + struct p9_fid { struct p9_client *clnt; u32 fid; diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index d2209ae9d18b..240e0de888c6 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -26,12 +26,40 @@ #ifndef NET_9P_TRANSPORT_H #define NET_9P_TRANSPORT_H +/** + * enum p9_trans_status - different states of underlying transports + * @Connected: transport is connected and healthy + * @Disconnected: transport has been disconnected + * @Hung: transport is connected by wedged + * + * This enumeration details the various states a transport + * instatiation can be in. + */ + enum p9_trans_status { Connected, Disconnected, Hung, }; +/** + * struct p9_trans - per-transport state and API + * @status: transport &p9_trans_status + * @msize: negotiated maximum packet size (duplicate from client) + * @extended: negotiated protocol extensions (duplicate from client) + * @priv: transport private data + * @close: member function to disconnect and close the transport + * @rpc: member function to issue a request to the transport + * + * This is the basic API for a transport instance. It is used as + * a handle by the client to issue requests. This interface is currently + * in flux during reorganization. + * + * Bugs: there is lots of duplicated data here and its not clear that + * the member functions need to be per-instance versus per transport + * module. + */ + struct p9_trans { enum p9_trans_status status; int msize; @@ -42,6 +70,21 @@ struct p9_trans { struct p9_fcall **rc); }; +/** + * struct p9_trans_module - transport module interface + * @list: used to maintain a list of currently available transports + * @name: the human-readable name of the transport + * @maxsize: transport provided maximum packet size + * @def: set if this transport should be considered the default + * @create: member function to create a new connection on this transport + * + * This is the basic API for a transport module which is registered by the + * transport module with the 9P core network module and used by the client + * to instantiate a new connection on a transport. + * + * Bugs: the transport module list isn't protected. + */ + struct p9_trans_module { struct list_head list; char *name; /* name of transport */ diff --git a/net/9p/conv.c b/net/9p/conv.c index 3fe35d532c87..44547201f5bc 100644 --- a/net/9p/conv.c +++ b/net/9p/conv.c @@ -197,7 +197,7 @@ static void buf_get_qid(struct cbuf *bufp, struct p9_qid *qid) /** * p9_size_wstat - calculate the size of a variable length stat struct - * @stat: metadata (stat) structure + * @wstat: metadata (stat) structure * @dotu: non-zero if 9P2000.u * */ @@ -511,6 +511,12 @@ p9_create_common(struct cbuf *bufp, u32 size, u8 id) return fc; } +/** + * p9_set_tag - set the tag field of an &p9_fcall structure + * @fc: fcall structure to set tag within + * @tag: tag id to set + */ + void p9_set_tag(struct p9_fcall *fc, u16 tag) { fc->tag = tag; @@ -518,6 +524,12 @@ void p9_set_tag(struct p9_fcall *fc, u16 tag) } EXPORT_SYMBOL(p9_set_tag); +/** + * p9_create_tversion - allocates and creates a T_VERSION request + * @msize: requested maximum data size + * @version: version string to negotiate + * + */ struct p9_fcall *p9_create_tversion(u32 msize, char *version) { int size; @@ -542,6 +554,16 @@ error: } EXPORT_SYMBOL(p9_create_tversion); +/** + * p9_create_tauth - allocates and creates a T_AUTH request + * @afid: handle to use for authentication protocol + * @uname: user name attempting to authenticate + * @aname: mount specifier for remote server + * @n_uname: numeric id for user attempting to authneticate + * @dotu: 9P2000.u extension flag + * + */ + struct p9_fcall *p9_create_tauth(u32 afid, char *uname, char *aname, u32 n_uname, int dotu) { @@ -580,6 +602,18 @@ error: } EXPORT_SYMBOL(p9_create_tauth); +/** + * p9_create_tattach - allocates and creates a T_ATTACH request + * @fid: handle to use for the new mount point + * @afid: handle to use for authentication protocol + * @uname: user name attempting to attach + * @aname: mount specifier for remote server + * @n_uname: numeric id for user attempting to attach + * @n_uname: numeric id for user attempting to attach + * @dotu: 9P2000.u extension flag + * + */ + struct p9_fcall * p9_create_tattach(u32 fid, u32 afid, char *uname, char *aname, u32 n_uname, int dotu) @@ -616,6 +650,12 @@ error: } EXPORT_SYMBOL(p9_create_tattach); +/** + * p9_create_tflush - allocates and creates a T_FLUSH request + * @oldtag: tag id for the transaction we are attempting to cancel + * + */ + struct p9_fcall *p9_create_tflush(u16 oldtag) { int size; @@ -639,6 +679,15 @@ error: } EXPORT_SYMBOL(p9_create_tflush); +/** + * p9_create_twalk - allocates and creates a T_FLUSH request + * @fid: handle we are traversing from + * @newfid: a new handle for this transaction + * @nwname: number of path elements to traverse + * @wnames: array of path elements + * + */ + struct p9_fcall *p9_create_twalk(u32 fid, u32 newfid, u16 nwname, char **wnames) { @@ -677,6 +726,13 @@ error: } EXPORT_SYMBOL(p9_create_twalk); +/** + * p9_create_topen - allocates and creates a T_OPEN request + * @fid: handle we are trying to open + * @mode: what mode we are trying to open the file in + * + */ + struct p9_fcall *p9_create_topen(u32 fid, u8 mode) { int size; @@ -701,6 +757,19 @@ error: } EXPORT_SYMBOL(p9_create_topen); +/** + * p9_create_tcreate - allocates and creates a T_CREATE request + * @fid: handle of directory we are trying to create in + * @name: name of the file we are trying to create + * @perm: permissions for the file we are trying to create + * @mode: what mode we are trying to open the file in + * @extension: 9p2000.u extension string (for special files) + * @dotu: 9p2000.u enabled flag + * + * Note: Plan 9 create semantics include opening the resulting file + * which is why mode is included. + */ + struct p9_fcall *p9_create_tcreate(u32 fid, char *name, u32 perm, u8 mode, char *extension, int dotu) { @@ -736,6 +805,13 @@ error: } EXPORT_SYMBOL(p9_create_tcreate); +/** + * p9_create_tread - allocates and creates a T_READ request + * @fid: handle of the file we are trying to read + * @offset: offset to start reading from + * @count: how many bytes to read + */ + struct p9_fcall *p9_create_tread(u32 fid, u64 offset, u32 count) { int size; @@ -761,6 +837,17 @@ error: } EXPORT_SYMBOL(p9_create_tread); +/** + * p9_create_twrite - allocates and creates a T_WRITE request from the kernel + * @fid: handle of the file we are trying to write + * @offset: offset to start writing at + * @count: how many bytes to write + * @data: data to write + * + * This function will create a requst with data buffers from the kernel + * such as the page cache. + */ + struct p9_fcall *p9_create_twrite(u32 fid, u64 offset, u32 count, const char *data) { @@ -794,6 +881,16 @@ error: } EXPORT_SYMBOL(p9_create_twrite); +/** + * p9_create_twrite_u - allocates and creates a T_WRITE request from userspace + * @fid: handle of the file we are trying to write + * @offset: offset to start writing at + * @count: how many bytes to write + * @data: data to write + * + * This function will create a request with data buffers from userspace + */ + struct p9_fcall *p9_create_twrite_u(u32 fid, u64 offset, u32 count, const char __user *data) { @@ -827,6 +924,14 @@ error: } EXPORT_SYMBOL(p9_create_twrite_u); +/** + * p9_create_tclunk - allocate a request to forget about a file handle + * @fid: handle of the file we closing or forgetting about + * + * clunk is used both to close open files and to discard transient handles + * which may be created during meta-data operations and hierarchy traversal. + */ + struct p9_fcall *p9_create_tclunk(u32 fid) { int size; @@ -850,6 +955,12 @@ error: } EXPORT_SYMBOL(p9_create_tclunk); +/** + * p9_create_tremove - allocate and create a request to remove a file + * @fid: handle of the file or directory we are removing + * + */ + struct p9_fcall *p9_create_tremove(u32 fid) { int size; @@ -873,6 +984,12 @@ error: } EXPORT_SYMBOL(p9_create_tremove); +/** + * p9_create_tstat - allocate and populate a request for attributes + * @fid: handle of the file or directory we are trying to get the attributes of + * + */ + struct p9_fcall *p9_create_tstat(u32 fid) { int size; @@ -896,6 +1013,14 @@ error: } EXPORT_SYMBOL(p9_create_tstat); +/** + * p9_create_tstat - allocate and populate a request to change attributes + * @fid: handle of the file or directory we are trying to change + * @wstat: &p9_stat structure with attributes we wish to set + * @dotu: 9p2000.u enabled flag + * + */ + struct p9_fcall *p9_create_twstat(u32 fid, struct p9_wstat *wstat, int dotu) { @@ -922,3 +1047,4 @@ error: return fc; } EXPORT_SYMBOL(p9_create_twstat); + diff --git a/net/9p/error.c b/net/9p/error.c index 64104b9cb422..388770c3631e 100644 --- a/net/9p/error.c +++ b/net/9p/error.c @@ -33,6 +33,13 @@ #include #include +/** + * struct errormap - map string errors from Plan 9 to Linux numeric ids + * @name: string sent over 9P + * @val: numeric id most closely representing @name + * @namelen: length of string + * @list: hash-table list for string lookup + */ struct errormap { char *name; int val; @@ -177,8 +184,7 @@ static struct errormap errmap[] = { }; /** - * p9_error_init - preload - * @errstr: error string + * p9_error_init - preload mappings into hash list * */ @@ -206,6 +212,7 @@ EXPORT_SYMBOL(p9_error_init); /** * errstr2errno - convert error string to error number * @errstr: error string + * @len: length of error string * */ diff --git a/net/9p/fcprint.c b/net/9p/fcprint.c index 40244fbd9b0d..53dd8e28dd8a 100644 --- a/net/9p/fcprint.c +++ b/net/9p/fcprint.c @@ -142,6 +142,14 @@ p9_printdata(char *buf, int buflen, u8 *data, int datalen) return p9_dumpdata(buf, buflen, data, datalen < 16?datalen:16); } +/** + * p9_printfcall - decode and print a protocol structure into a buffer + * @buf: buffer to deposit decoded structure into + * @buflen: available space in buffer + * @fc: protocol rpc structure of type &p9_fcall + * @extended: whether or not session is operating with extended protocol + */ + int p9_printfcall(char *buf, int buflen, struct p9_fcall *fc, int extended) { diff --git a/net/9p/mod.c b/net/9p/mod.c index c285aab2af04..c6d9695949e0 100644 --- a/net/9p/mod.c +++ b/net/9p/mod.c @@ -39,9 +39,6 @@ module_param_named(debug, p9_debug_level, uint, 0); MODULE_PARM_DESC(debug, "9P debugging level"); #endif -extern int p9_mux_global_init(void); -extern void p9_mux_global_exit(void); - /* * Dynamic Transport Registration Routines * @@ -52,7 +49,7 @@ static struct p9_trans_module *v9fs_default_transport; /** * v9fs_register_trans - register a new transport with 9p - * @m - structure describing the transport module and entry points + * @m: structure describing the transport module and entry points * */ void v9fs_register_trans(struct p9_trans_module *m) @@ -65,7 +62,7 @@ EXPORT_SYMBOL(v9fs_register_trans); /** * v9fs_match_trans - match transport versus registered transports - * @arg: string identifying transport + * @name: string identifying transport * */ struct p9_trans_module *v9fs_match_trans(const substring_t *name) diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index f624dff76852..c6eda999fa7d 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -47,12 +47,29 @@ #define SCHED_TIMEOUT 10 #define MAXPOLLWADDR 2 +/** + * struct p9_fd_opts - per-transport options + * @rfd: file descriptor for reading (trans=fd) + * @wfd: file descriptor for writing (trans=fd) + * @port: port to connect to (trans=tcp) + * + */ + struct p9_fd_opts { int rfd; int wfd; u16 port; }; + +/** + * struct p9_trans_fd - transport state + * @rd: reference to file to read from + * @wr: reference of file to write to + * @conn: connection state reference + * + */ + struct p9_trans_fd { struct file *rd; struct file *wr; @@ -90,10 +107,24 @@ enum { }; struct p9_req; - typedef void (*p9_conn_req_callback)(struct p9_req *req, void *a); + +/** + * struct p9_req - fd mux encoding of an rpc transaction + * @lock: protects req_list + * @tag: numeric tag for rpc transaction + * @tcall: request &p9_fcall structure + * @rcall: response &p9_fcall structure + * @err: error state + * @cb: callback for when response is received + * @cba: argument to pass to callback + * @flush: flag to indicate RPC has been flushed + * @req_list: list link for higher level objects to chain requests + * + */ + struct p9_req { - spinlock_t lock; /* protect request structure */ + spinlock_t lock; int tag; struct p9_fcall *tcall; struct p9_fcall *rcall; @@ -104,7 +135,39 @@ struct p9_req { struct list_head req_list; }; -struct p9_mux_poll_task; +struct p9_mux_poll_task { + struct task_struct *task; + struct list_head mux_list; + int muxnum; +}; + +/** + * struct p9_conn - fd mux connection state information + * @lock: protects mux_list (?) + * @mux_list: list link for mux to manage multiple connections (?) + * @poll_task: task polling on this connection + * @msize: maximum size for connection (dup) + * @extended: 9p2000.u flag (dup) + * @trans: reference to transport instance for this connection + * @tagpool: id accounting for transactions + * @err: error state + * @equeue: event wait_q (?) + * @req_list: accounting for requests which have been sent + * @unsent_req_list: accounting for requests that haven't been sent + * @rcall: current response &p9_fcall structure + * @rpos: read position in current frame + * @rbuf: current read buffer + * @wpos: write position for current frame + * @wsize: amount of data to write for current frame + * @wbuf: current write buffer + * @poll_wait: array of wait_q's for various worker threads + * @poll_waddr: ???? + * @pt: poll state + * @rq: current read work + * @wq: current write work + * @wsched: ???? + * + */ struct p9_conn { spinlock_t lock; /* protect lock structure */ @@ -132,11 +195,16 @@ struct p9_conn { unsigned long wsched; }; -struct p9_mux_poll_task { - struct task_struct *task; - struct list_head mux_list; - int muxnum; -}; +/** + * struct p9_mux_rpc - fd mux rpc accounting structure + * @m: connection this request was issued on + * @err: error state + * @tcall: request &p9_fcall + * @rcall: response &p9_fcall + * @wqueue: wait queue that client is blocked on for this rpc + * + * Bug: isn't this information duplicated elsewhere like &p9_req + */ struct p9_mux_rpc { struct p9_conn *m; @@ -207,10 +275,12 @@ static void p9_mux_put_tag(struct p9_conn *m, u16 tag) /** * p9_mux_calc_poll_procs - calculates the number of polling procs - * based on the number of mounted v9fs filesystems. + * @muxnum: number of mounts * + * Calculation is based on the number of mounted v9fs filesystems. * The current implementation returns sqrt of the number of mounts. */ + static int p9_mux_calc_poll_procs(int muxnum) { int n; @@ -331,12 +401,11 @@ static void p9_mux_poll_stop(struct p9_conn *m) /** * p9_conn_create - allocate and initialize the per-session mux data - * Creates the polling task if this is the first session. + * @trans: transport structure * - * @trans - transport structure - * @msize - maximum message size - * @extended - extended flag + * Note: Creates the polling task if this is the first session. */ + static struct p9_conn *p9_conn_create(struct p9_trans *trans) { int i, n; @@ -406,7 +475,10 @@ static struct p9_conn *p9_conn_create(struct p9_trans *trans) /** * p9_mux_destroy - cancels all pending requests and frees mux resources + * @m: mux to destroy + * */ + static void p9_conn_destroy(struct p9_conn *m) { P9_DPRINTK(P9_DEBUG_MUX, "mux %p prev %p next %p\n", m, @@ -429,9 +501,14 @@ static void p9_conn_destroy(struct p9_conn *m) } /** - * p9_pollwait - called by files poll operation to add v9fs-poll task - * to files wait queue + * p9_pollwait - add poll task to the wait queue + * @filp: file pointer being polled + * @wait_address: wait_q to block on + * @p: poll state + * + * called by files poll operation to add v9fs-poll task to files wait queue */ + static void p9_pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) { @@ -462,7 +539,10 @@ p9_pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) /** * p9_poll_mux - polls a mux and schedules read or write works if necessary + * @m: connection to poll + * */ + static void p9_poll_mux(struct p9_conn *m) { int n; @@ -499,9 +579,14 @@ static void p9_poll_mux(struct p9_conn *m) } /** - * p9_poll_proc - polls all v9fs transports for new events and queues - * the appropriate work to the work queue + * p9_poll_proc - poll worker thread + * @a: thread state and arguments + * + * polls all v9fs transports for new events and queues the appropriate + * work to the work queue + * */ + static int p9_poll_proc(void *a) { struct p9_conn *m, *mtmp; @@ -527,7 +612,10 @@ static int p9_poll_proc(void *a) /** * p9_write_work - called when a transport can send some data + * @work: container for work to be done + * */ + static void p9_write_work(struct work_struct *work) { int n, err; @@ -638,7 +726,10 @@ static void process_request(struct p9_conn *m, struct p9_req *req) /** * p9_read_work - called when there is some data to be read from a transport + * @work: container of work to be done + * */ + static void p9_read_work(struct work_struct *work) { int n, err; @@ -793,7 +884,9 @@ error: * @tc: request to be sent * @cb: callback function to call when response is received * @cba: parameter to pass to the callback function + * */ + static struct p9_req *p9_send_request(struct p9_conn *m, struct p9_fcall *tc, p9_conn_req_callback cb, void *cba) @@ -961,10 +1054,12 @@ p9_conn_rpc_cb(struct p9_req *req, void *a) /** * p9_fd_rpc- sends 9P request and waits until a response is available. * The function can be interrupted. - * @m: mux data + * @t: transport data * @tc: request to be sent * @rc: pointer where a pointer to the response is stored + * */ + int p9_fd_rpc(struct p9_trans *t, struct p9_fcall *tc, struct p9_fcall **rc) { @@ -1041,8 +1136,10 @@ p9_fd_rpc(struct p9_trans *t, struct p9_fcall *tc, struct p9_fcall **rc) * @m: mux data * @tc: request to be sent * @cb: callback function to be called when response arrives - * @cba: value to pass to the callback function + * @a: value to pass to the callback function + * */ + int p9_conn_rpcnb(struct p9_conn *m, struct p9_fcall *tc, p9_conn_req_callback cb, void *a) { @@ -1065,7 +1162,9 @@ int p9_conn_rpcnb(struct p9_conn *m, struct p9_fcall *tc, * p9_conn_cancel - cancel all pending requests with error * @m: mux data * @err: error code + * */ + void p9_conn_cancel(struct p9_conn *m, int err) { struct p9_req *req, *rtmp; @@ -1099,7 +1198,7 @@ void p9_conn_cancel(struct p9_conn *m, int err) /** * v9fs_parse_options - parse mount options into session structure * @options: options string passed from mount - * @v9ses: existing v9fs session information + * @opts: transport-specific structure to parse options into * */ @@ -1193,11 +1292,12 @@ static int p9_socket_open(struct p9_trans *trans, struct socket *csocket) /** * p9_fd_read- read from a fd - * @v9ses: session information + * @trans: transport instance state * @v: buffer to receive data into * @len: size of receive buffer * */ + static int p9_fd_read(struct p9_trans *trans, void *v, int len) { int ret; @@ -1220,11 +1320,12 @@ static int p9_fd_read(struct p9_trans *trans, void *v, int len) /** * p9_fd_write - write to a socket - * @v9ses: session information + * @trans: transport instance state * @v: buffer to send data from * @len: size of send buffer * */ + static int p9_fd_write(struct p9_trans *trans, void *v, int len) { int ret; @@ -1296,6 +1397,7 @@ end: * @trans: private socket structure * */ + static void p9_fd_close(struct p9_trans *trans) { struct p9_trans_fd *ts; diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index de7a9f532edc..0bab1f23590e 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -55,23 +55,69 @@ static int chan_index; #define P9_INIT_MAXTAG 16 -#define REQ_STATUS_IDLE 0 -#define REQ_STATUS_SENT 1 -#define REQ_STATUS_RCVD 2 -#define REQ_STATUS_FLSH 3 + +/** + * enum p9_req_status_t - virtio request status + * @REQ_STATUS_IDLE: request slot unused + * @REQ_STATUS_SENT: request sent to server + * @REQ_STATUS_RCVD: response received from server + * @REQ_STATUS_FLSH: request has been flushed + * + * The @REQ_STATUS_IDLE state is used to mark a request slot as unused + * but use is actually tracked by the idpool structure which handles tag + * id allocation. + * + */ + +enum p9_req_status_t { + REQ_STATUS_IDLE, + REQ_STATUS_SENT, + REQ_STATUS_RCVD, + REQ_STATUS_FLSH, +}; + +/** + * struct p9_req_t - virtio request slots + * @status: status of this request slot + * @wq: wait_queue for the client to block on for this request + * + * The virtio transport uses an array to track outstanding requests + * instead of a list. While this may incurr overhead during initial + * allocation or expansion, it makes request lookup much easier as the + * tag id is a index into an array. (We use tag+1 so that we can accomodate + * the -1 tag for the T_VERSION request). + * This also has the nice effect of only having to allocate wait_queues + * once, instead of constantly allocating and freeing them. Its possible + * other resources could benefit from this scheme as well. + * + */ struct p9_req_t { int status; wait_queue_head_t *wq; }; -/* We keep all per-channel information in a structure. +/** + * struct virtio_chan - per-instance transport information + * @initialized: whether the channel is initialized + * @inuse: whether the channel is in use + * @lock: protects multiple elements within this structure + * @vdev: virtio dev associated with this channel + * @vq: virtio queue associated with this channel + * @tagpool: accounting for tag ids (and request slots) + * @reqs: array of request slots + * @max_tag: current number of request_slots allocated + * @sg: scatter gather list which is used to pack a request (protected?) + * + * We keep all per-channel information in a structure. * This structure is allocated within the devices dev->mem space. * A pointer to the structure will get put in the transport private. + * */ + static struct virtio_chan { - bool initialized; /* channel is initialized */ - bool inuse; /* channel is in use */ + bool initialized; + bool inuse; spinlock_t lock; @@ -86,7 +132,19 @@ static struct virtio_chan { struct scatterlist sg[VIRTQUEUE_NUM]; } channels[MAX_9P_CHAN]; -/* Lookup requests by tag */ +/** + * p9_lookup_tag - Lookup requests by tag + * @c: virtio channel to lookup tag within + * @tag: numeric id for transaction + * + * this is a simple array lookup, but will grow the + * request_slots as necessary to accomodate transaction + * ids which did not previously have a slot. + * + * Bugs: there is currently no upper limit on request slots set + * here, but that should be constrained by the id accounting. + */ + static struct p9_req_t *p9_lookup_tag(struct virtio_chan *c, u16 tag) { /* This looks up the original request by tag so we know which @@ -130,6 +188,15 @@ static unsigned int rest_of_page(void *data) return PAGE_SIZE - ((unsigned long)data % PAGE_SIZE); } +/** + * p9_virtio_close - reclaim resources of a channel + * @trans: transport state + * + * This reclaims a channel by freeing its resources and + * reseting its inuse flag. + * + */ + static void p9_virtio_close(struct p9_trans *trans) { struct virtio_chan *chan = trans->priv; @@ -151,6 +218,19 @@ static void p9_virtio_close(struct p9_trans *trans) kfree(trans); } +/** + * req_done - callback which signals activity from the server + * @vq: virtio queue activity was received on + * + * This notifies us that the server has triggered some activity + * on the virtio channel - most likely a response to request we + * sent. Figure out which requests now have responses and wake up + * those threads. + * + * Bugs: could do with some additional sanity checking, but appears to work. + * + */ + static void req_done(struct virtqueue *vq) { struct virtio_chan *chan = vq->vdev->priv; @@ -169,6 +249,20 @@ static void req_done(struct virtqueue *vq) spin_unlock_irqrestore(&chan->lock, flags); } +/** + * pack_sg_list - pack a scatter gather list from a linear buffer + * @sg: scatter/gather list to pack into + * @start: which segment of the sg_list to start at + * @limit: maximum segment to pack data to + * @data: data to pack into scatter/gather list + * @count: amount of data to pack into the scatter/gather list + * + * sg_lists have multiple segments of various sizes. This will pack + * arbitrary data into an existing scatter gather list, segmenting the + * data as necessary within constraints. + * + */ + static int pack_sg_list(struct scatterlist *sg, int start, int limit, char *data, int count) @@ -189,6 +283,14 @@ pack_sg_list(struct scatterlist *sg, int start, int limit, char *data, return index-start; } +/** + * p9_virtio_rpc - issue a request and wait for a response + * @t: transport state + * @tc: &p9_fcall request to transmit + * @rc: &p9_fcall to put reponse into + * + */ + static int p9_virtio_rpc(struct p9_trans *t, struct p9_fcall *tc, struct p9_fcall **rc) { @@ -263,6 +365,16 @@ p9_virtio_rpc(struct p9_trans *t, struct p9_fcall *tc, struct p9_fcall **rc) return 0; } +/** + * p9_virtio_probe - probe for existence of 9P virtio channels + * @vdev: virtio device to probe + * + * This probes for existing virtio channels. At present only + * a single channel is in use, so in the future more work may need + * to be done here. + * + */ + static int p9_virtio_probe(struct virtio_device *vdev) { int err; @@ -307,11 +419,28 @@ fail: return err; } -/* This sets up a transport channel for 9p communication. Right now + +/** + * p9_virtio_create - allocate a new virtio channel + * @devname: string identifying the channel to connect to (unused) + * @args: args passed from sys_mount() for per-transport options (unused) + * @msize: requested maximum packet size + * @extended: 9p2000.u enabled flag + * + * This sets up a transport channel for 9p communication. Right now * we only match the first available channel, but eventually we couldlook up * alternate channels by matching devname versus a virtio_config entry. * We use a simple reference count mechanism to ensure that only a single - * mount has a channel open at a time. */ + * mount has a channel open at a time. + * + * Bugs: doesn't allow identification of a specific channel + * to allocate, channels are allocated sequentially. This was + * a pragmatic decision to get things rolling, but ideally some + * way of identifying the channel to attach to would be nice + * if we are going to support multiple channels. + * + */ + static struct p9_trans * p9_virtio_create(const char *devname, char *args, int msize, unsigned char extended) @@ -360,6 +489,12 @@ p9_virtio_create(const char *devname, char *args, int msize, return trans; } +/** + * p9_virtio_remove - clean up resources associated with a virtio device + * @vdev: virtio device to remove + * + */ + static void p9_virtio_remove(struct virtio_device *vdev) { struct virtio_chan *chan = vdev->priv; diff --git a/net/9p/util.c b/net/9p/util.c index ef7215565d88..4d5646045330 100644 --- a/net/9p/util.c +++ b/net/9p/util.c @@ -32,11 +32,23 @@ #include #include +/** + * struct p9_idpool - per-connection accounting for tag idpool + * @lock: protects the pool + * @pool: idr to allocate tag id from + * + */ + struct p9_idpool { spinlock_t lock; struct idr pool; }; +/** + * p9_idpool_create - create a new per-connection id pool + * + */ + struct p9_idpool *p9_idpool_create(void) { struct p9_idpool *p; @@ -52,6 +64,11 @@ struct p9_idpool *p9_idpool_create(void) } EXPORT_SYMBOL(p9_idpool_create); +/** + * p9_idpool_destroy - create a new per-connection id pool + * @p: idpool to destory + */ + void p9_idpool_destroy(struct p9_idpool *p) { idr_destroy(&p->pool); @@ -61,9 +78,9 @@ EXPORT_SYMBOL(p9_idpool_destroy); /** * p9_idpool_get - allocate numeric id from pool - * @p - pool to allocate from + * @p: pool to allocate from * - * XXX - This seems to be an awful generic function, should it be in idr.c with + * Bugs: This seems to be an awful generic function, should it be in idr.c with * the lock included in struct idr? */ @@ -94,9 +111,10 @@ EXPORT_SYMBOL(p9_idpool_get); /** * p9_idpool_put - release numeric id from pool - * @p - pool to allocate from + * @id: numeric id which is being released + * @p: pool to release id into * - * XXX - This seems to be an awful generic function, should it be in idr.c with + * Bugs: This seems to be an awful generic function, should it be in idr.c with * the lock included in struct idr? */ @@ -111,11 +129,13 @@ EXPORT_SYMBOL(p9_idpool_put); /** * p9_idpool_check - check if the specified id is available - * @id - id to check - * @p - pool + * @id: id to check + * @p: pool to check */ + int p9_idpool_check(int id, struct p9_idpool *p) { return idr_find(&p->pool, id) != NULL; } EXPORT_SYMBOL(p9_idpool_check); + -- cgit From 887b3ece65be7b643dfdae0d433c91a26a3f437d Mon Sep 17 00:00:00 2001 From: Eric Van Hensbergen Date: Thu, 8 May 2008 20:26:37 -0500 Subject: 9p: fix error path during early mount There was some cleanup issues during early mount which would trigger a kernel bug for certain types of failure. This patch reorganizes the cleanup to get rid of the bad behavior. This also merges the 9pnet and 9pnet_fd modules for the purpose of configuration and initialization. Keeping the fd transport separate from the core 9pnet code seemed like a good idea at the time, but in practice has caused more harm and confusion than good. Signed-off-by: Eric Van Hensbergen --- fs/9p/v9fs.c | 13 +++++++------ fs/9p/vfs_super.c | 34 ++++++++++++++++------------------ include/net/9p/9p.h | 1 + include/net/9p/transport.h | 1 - net/9p/Kconfig | 10 ---------- net/9p/Makefile | 3 --- net/9p/mod.c | 1 + net/9p/trans_fd.c | 29 ++++++++++++++++++++++------- 8 files changed, 47 insertions(+), 45 deletions(-) (limited to 'include') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 5c1ccaf0416c..047c791427aa 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -206,12 +206,14 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->uid = ~0; v9ses->dfltuid = V9FS_DEFUID; v9ses->dfltgid = V9FS_DEFGID; - v9ses->options = kstrdup(data, GFP_KERNEL); - if (!v9ses->options) { - P9_DPRINTK(P9_DEBUG_ERROR, + if (data) { + v9ses->options = kstrdup(data, GFP_KERNEL); + if (!v9ses->options) { + P9_DPRINTK(P9_DEBUG_ERROR, "failed to allocate copy of option string\n"); - retval = -ENOMEM; - goto error; + retval = -ENOMEM; + goto error; + } } rc = v9fs_parse_options(v9ses); @@ -260,7 +262,6 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, return fid; error: - v9fs_session_close(v9ses); return ERR_PTR(retval); } diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index ba10f172626d..bf59c3960494 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -128,29 +128,26 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, fid = v9fs_session_init(v9ses, dev_name, data); if (IS_ERR(fid)) { retval = PTR_ERR(fid); - fid = NULL; - kfree(v9ses); - v9ses = NULL; - goto error; + goto close_session; } st = p9_client_stat(fid); if (IS_ERR(st)) { retval = PTR_ERR(st); - goto error; + goto clunk_fid; } sb = sget(fs_type, NULL, v9fs_set_super, v9ses); if (IS_ERR(sb)) { retval = PTR_ERR(sb); - goto error; + goto free_stat; } v9fs_fill_super(sb, v9ses, flags); inode = v9fs_get_inode(sb, S_IFDIR | mode); if (IS_ERR(inode)) { retval = PTR_ERR(inode); - goto error; + goto release_sb; } inode->i_uid = uid; @@ -159,7 +156,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, root = d_alloc_root(inode); if (!root) { retval = -ENOMEM; - goto error; + goto release_sb; } sb->s_root = root; @@ -170,21 +167,22 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, return simple_set_mnt(mnt, sb); -error: - kfree(st); - if (fid) - p9_client_clunk(fid); - - if (v9ses) { - v9fs_session_close(v9ses); - kfree(v9ses); - } - +release_sb: if (sb) { up_write(&sb->s_umount); deactivate_super(sb); } +free_stat: + kfree(st); + +clunk_fid: + p9_client_clunk(fid); + +close_session: + v9fs_session_close(v9ses); + kfree(v9ses); + return retval; } diff --git a/include/net/9p/9p.h b/include/net/9p/9p.h index 7bfb2f2e423c..b3d3e27c6299 100644 --- a/include/net/9p/9p.h +++ b/include/net/9p/9p.h @@ -595,4 +595,5 @@ int p9_idpool_check(int id, struct p9_idpool *p); int p9_error_init(void); int p9_errstr2errno(char *, int); +int p9_trans_fd_init(void); #endif /* NET_9P_H */ diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index 240e0de888c6..0db3a4038dc0 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -96,5 +96,4 @@ struct p9_trans_module { void v9fs_register_trans(struct p9_trans_module *m); struct p9_trans_module *v9fs_match_trans(const substring_t *name); struct p9_trans_module *v9fs_default_trans(void); - #endif /* NET_9P_TRANSPORT_H */ diff --git a/net/9p/Kconfig b/net/9p/Kconfig index bafc50c9e6ff..ff34c5acc130 100644 --- a/net/9p/Kconfig +++ b/net/9p/Kconfig @@ -13,16 +13,6 @@ menuconfig NET_9P If unsure, say N. -config NET_9P_FD - depends on NET_9P - default y if NET_9P - tristate "9P File Descriptor Transports (Experimental)" - help - This builds support for file descriptor transports for 9p - which includes support for TCP/IP, named pipes, or passed - file descriptors. TCP/IP is the default transport for 9p, - so if you are going to use 9p, you'll likely want this. - config NET_9P_VIRTIO depends on NET_9P && EXPERIMENTAL && VIRTIO tristate "9P Virtio Transport (Experimental)" diff --git a/net/9p/Makefile b/net/9p/Makefile index 8a1051101898..519219480db1 100644 --- a/net/9p/Makefile +++ b/net/9p/Makefile @@ -1,5 +1,4 @@ obj-$(CONFIG_NET_9P) := 9pnet.o -obj-$(CONFIG_NET_9P_FD) += 9pnet_fd.o obj-$(CONFIG_NET_9P_VIRTIO) += 9pnet_virtio.o 9pnet-objs := \ @@ -9,8 +8,6 @@ obj-$(CONFIG_NET_9P_VIRTIO) += 9pnet_virtio.o error.o \ fcprint.o \ util.o \ - -9pnet_fd-objs := \ trans_fd.o \ 9pnet_virtio-objs := \ diff --git a/net/9p/mod.c b/net/9p/mod.c index c6d9695949e0..bdee1fb7cc62 100644 --- a/net/9p/mod.c +++ b/net/9p/mod.c @@ -107,6 +107,7 @@ static int __init init_p9(void) p9_error_init(); printk(KERN_INFO "Installing 9P2000 support\n"); + p9_trans_fd_init(); return ret; } diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index 97b103b70499..4507f744f44e 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -1433,6 +1433,23 @@ static void p9_fd_close(struct p9_trans *trans) kfree(ts); } +/* + * stolen from NFS - maybe should be made a generic function? + */ +static inline int valid_ipaddr4(const char *buf) +{ + int rc, count, in[4]; + + rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]); + if (rc != 4) + return -EINVAL; + for (count = 0; count < 4; count++) { + if (in[count] > 255) + return -EINVAL; + } + return 0; +} + static struct p9_trans * p9_trans_create_tcp(const char *addr, char *args, int msize, unsigned char dotu) { @@ -1447,6 +1464,9 @@ p9_trans_create_tcp(const char *addr, char *args, int msize, unsigned char dotu) if (err < 0) return ERR_PTR(err); + if (valid_ipaddr4(addr) < 0) + return ERR_PTR(-EINVAL); + csocket = NULL; trans = kmalloc(sizeof(struct p9_trans), GFP_KERNEL); if (!trans) @@ -1625,7 +1645,7 @@ static struct p9_trans_module p9_fd_trans = { .create = p9_trans_create_fd, }; -static int __init p9_trans_fd_init(void) +int p9_trans_fd_init(void) { int ret = p9_mux_global_init(); if (ret) { @@ -1639,9 +1659,4 @@ static int __init p9_trans_fd_init(void) return 0; } - -module_init(p9_trans_fd_init); - -MODULE_AUTHOR("Latchesar Ionkov "); -MODULE_AUTHOR("Eric Van Hensbergen "); -MODULE_LICENSE("GPL"); +EXPORT_SYMBOL(p9_trans_fd_init); -- cgit From b7cffc1f29c1bc729bc50c863c87f93f9b70994b Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Wed, 14 May 2008 16:05:42 -0700 Subject: asm-{alpha,h8300,um,v850,xtensa}/param.h: unbreak HZ for userspace I noticed this because alpha was broken due to the recent commit commit bdc807871d58285737d50dc6163d0feb72cb0dc2 ("avoid overflows in kernel/time.c"). Most arches do something like this in their asm/param.h: #ifdef __KERNEL__ # define HZ CONFIG_HZ #else # define HZ 100 #endif A few arches though (namely alpha/h8300/um/v850/xtensa) either do no set HZ at all for !__KERNEL__, or they set it wrongly. This should bring all arches in line by setting up HZ for userspace. Without this currently perl 5.10 doesn't build on alpha: perl.c: In function 'perl_construct': perl.c:388: error: 'CONFIG_HZ' undeclared (first use in this function) -> http://buildd.debian.org/fetch.cgi?pkg=perl;ver=5.10.0-10;arch=alpha;stamp=1210252894 Signed-off-by: Mike Frysinger Cc: Richard Henderson Cc: Ivan Kokshaysky Cc: Yoshinori Sato Cc: Jeff Dike Cc: Chris Zankel Cc: maximilian attems Signed-off-by: Andrew Morton [ HZ on alpha is 1024 for historical reasons. - Linus ] Signed-off-by: Linus Torvalds --- include/asm-alpha/param.h | 4 ++++ include/asm-h8300/param.h | 8 +++----- include/asm-um/param.h | 2 ++ include/asm-v850/param.h | 2 ++ include/asm-xtensa/param.h | 2 ++ 5 files changed, 13 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/asm-alpha/param.h b/include/asm-alpha/param.h index 0982f1d39499..e691ecfedb2c 100644 --- a/include/asm-alpha/param.h +++ b/include/asm-alpha/param.h @@ -5,8 +5,12 @@ hardware ignores reprogramming. We also need userland buy-in to the change in HZ, since this is visible in the wait4 resources etc. */ +#ifdef __KERNEL__ #define HZ CONFIG_HZ #define USER_HZ HZ +#else +#define HZ 1024 +#endif #define EXEC_PAGESIZE 8192 diff --git a/include/asm-h8300/param.h b/include/asm-h8300/param.h index 04f64f100379..1c72fb8080ff 100644 --- a/include/asm-h8300/param.h +++ b/include/asm-h8300/param.h @@ -1,14 +1,12 @@ #ifndef _H8300_PARAM_H #define _H8300_PARAM_H - -#ifndef HZ -#define HZ CONFIG_HZ -#endif - #ifdef __KERNEL__ +#define HZ CONFIG_HZ #define USER_HZ HZ #define CLOCKS_PER_SEC (USER_HZ) +#else +#define HZ 100 #endif #define EXEC_PAGESIZE 4096 diff --git a/include/asm-um/param.h b/include/asm-um/param.h index 4cd4a226f8c1..e44f4e60d16d 100644 --- a/include/asm-um/param.h +++ b/include/asm-um/param.h @@ -13,6 +13,8 @@ #define HZ CONFIG_HZ #define USER_HZ 100 /* .. some user interfaces are in "ticks" */ #define CLOCKS_PER_SEC (USER_HZ) /* frequency at which times() counts */ +#else +#define HZ 100 #endif #endif diff --git a/include/asm-v850/param.h b/include/asm-v850/param.h index 281832690290..4391f5fe0204 100644 --- a/include/asm-v850/param.h +++ b/include/asm-v850/param.h @@ -26,6 +26,8 @@ # define HZ CONFIG_HZ # define USER_HZ 100 # define CLOCKS_PER_SEC USER_HZ +#else +# define HZ 100 #endif #endif /* __V850_PARAM_H__ */ diff --git a/include/asm-xtensa/param.h b/include/asm-xtensa/param.h index 82ad34d92d35..ba03d5aeab6b 100644 --- a/include/asm-xtensa/param.h +++ b/include/asm-xtensa/param.h @@ -15,6 +15,8 @@ # define HZ CONFIG_HZ /* internal timer frequency */ # define USER_HZ 100 /* for user interfaces in "ticks" */ # define CLOCKS_PER_SEC (USER_HZ) /* frequnzy at which times() counts */ +#else +# define HZ 100 #endif #define EXEC_PAGESIZE 4096 -- cgit From 3fc957721d18c93662f7d4dab455b80f53dd2641 Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Wed, 14 May 2008 16:05:49 -0700 Subject: lib: create common ascii hex array Add a common hex array in hexdump.c so everyone can use it. Add a common hi/lo helper to avoid the shifting masking that is done to get the upper and lower nibbles of a byte value. Pull the pack_hex_byte helper from kgdb as it is opencoded many places in the tree that will be consolidated. Signed-off-by: Harvey Harrison Acked-by: Paul Mundt Cc: Jason Wessel Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/sh/kernel/kgdb_stub.c | 8 -------- drivers/pnp/support.c | 8 ++++---- include/linux/kernel.h | 12 +++++++++++- kernel/kgdb.c | 8 -------- lib/hexdump.c | 7 +++++-- 5 files changed, 20 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/arch/sh/kernel/kgdb_stub.c b/arch/sh/kernel/kgdb_stub.c index d453c3a1c79f..832641bbd47d 100644 --- a/arch/sh/kernel/kgdb_stub.c +++ b/arch/sh/kernel/kgdb_stub.c @@ -330,14 +330,6 @@ static char *ebin_to_mem(const char *buf, char *mem, int count) return mem; } -/* Pack a hex byte */ -static char *pack_hex_byte(char *pkt, int byte) -{ - *pkt++ = hexchars[(byte >> 4) & 0xf]; - *pkt++ = hexchars[(byte & 0xf)]; - return pkt; -} - /* Scan for the start char '$', read the packet and check the checksum */ static void get_packet(char *buffer, int buflen) { diff --git a/drivers/pnp/support.c b/drivers/pnp/support.c index 3eba85ed729c..95b076c18c07 100644 --- a/drivers/pnp/support.c +++ b/drivers/pnp/support.c @@ -45,10 +45,10 @@ void pnp_eisa_id_to_string(u32 id, char *str) str[0] = 'A' + ((id >> 26) & 0x3f) - 1; str[1] = 'A' + ((id >> 21) & 0x1f) - 1; str[2] = 'A' + ((id >> 16) & 0x1f) - 1; - str[3] = hex_asc((id >> 12) & 0xf); - str[4] = hex_asc((id >> 8) & 0xf); - str[5] = hex_asc((id >> 4) & 0xf); - str[6] = hex_asc((id >> 0) & 0xf); + str[3] = hex_asc_hi(id >> 8); + str[4] = hex_asc_lo(id >> 8); + str[5] = hex_asc_hi(id); + str[6] = hex_asc_lo(id); str[7] = '\0'; } diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 4d46e299afb5..792bf0aa779b 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -276,7 +276,17 @@ extern void print_hex_dump(const char *level, const char *prefix_str, const void *buf, size_t len, bool ascii); extern void print_hex_dump_bytes(const char *prefix_str, int prefix_type, const void *buf, size_t len); -#define hex_asc(x) "0123456789abcdef"[x] + +extern const char hex_asc[]; +#define hex_asc_lo(x) hex_asc[((x) & 0x0f)] +#define hex_asc_hi(x) hex_asc[((x) & 0xf0) >> 4] + +static inline char *pack_hex_byte(char *buf, u8 byte) +{ + *buf++ = hex_asc_hi(byte); + *buf++ = hex_asc_lo(byte); + return buf; +} #define pr_emerg(fmt, arg...) \ printk(KERN_EMERG fmt, ##arg) diff --git a/kernel/kgdb.c b/kernel/kgdb.c index 39e31a036f5b..14787de568b3 100644 --- a/kernel/kgdb.c +++ b/kernel/kgdb.c @@ -346,14 +346,6 @@ static void put_packet(char *buffer) } } -static char *pack_hex_byte(char *pkt, u8 byte) -{ - *pkt++ = hexchars[byte >> 4]; - *pkt++ = hexchars[byte & 0xf]; - - return pkt; -} - /* * Convert the memory pointed to by mem into hex, placing result in buf. * Return a pointer to the last char put in buf (null). May return an error. diff --git a/lib/hexdump.c b/lib/hexdump.c index 343546550dc9..f07c0db81d26 100644 --- a/lib/hexdump.c +++ b/lib/hexdump.c @@ -12,6 +12,9 @@ #include #include +const char hex_asc[] = "0123456789abcdef"; +EXPORT_SYMBOL(hex_asc); + /** * hex_dump_to_buffer - convert a blob of data to "hex ASCII" in memory * @buf: data blob to dump @@ -93,8 +96,8 @@ void hex_dump_to_buffer(const void *buf, size_t len, int rowsize, for (j = 0; (j < rowsize) && (j < len) && (lx + 4) < linebuflen; j++) { ch = ptr[j]; - linebuf[lx++] = hex_asc(ch >> 4); - linebuf[lx++] = hex_asc(ch & 0x0f); + linebuf[lx++] = hex_asc_hi(ch); + linebuf[lx++] = hex_asc_lo(ch); linebuf[lx++] = ' '; } ascii_column = 3 * rowsize + 2; -- cgit From 44c81433e8b05dbc85985d939046f10f95901184 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 14 May 2008 16:05:51 -0700 Subject: per_cpu: fix DEFINE_PER_CPU_SHARED_ALIGNED for modules Current module loader lookups ".data.percpu" ELF section to perform per_cpu relocation. But DEFINE_PER_CPU_SHARED_ALIGNED() uses another section (".data.percpu.shared_aligned"), currently only handled in vmlinux.lds, not by module loader. To correct this problem, instead of adding logic into module loader, or using at build time a module.lds file for all arches to group ".data.percpu.shared_aligned" into ".data.percpu", just use ".data.percpu" for modules. Alignment requirements are correctly handled by ld and module loader. Signed-off-by: Eric Dumazet Cc: Rusty Russell Cc: Fenghua Yu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/percpu.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/percpu.h b/include/linux/percpu.h index d746a2abb322..4cdd393e71e1 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -13,8 +13,14 @@ __attribute__((__section__(".data.percpu"))) \ PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name +#ifdef MODULE +#define SHARED_ALIGNED_SECTION ".data.percpu" +#else +#define SHARED_ALIGNED_SECTION ".data.percpu.shared_aligned" +#endif + #define DEFINE_PER_CPU_SHARED_ALIGNED(type, name) \ - __attribute__((__section__(".data.percpu.shared_aligned"))) \ + __attribute__((__section__(SHARED_ALIGNED_SECTION))) \ PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name \ ____cacheline_aligned_in_smp #else -- cgit From 1c12c4cf9411eb130b245fa8d0fbbaf989477c7b Mon Sep 17 00:00:00 2001 From: Venki Pallipadi Date: Wed, 14 May 2008 16:05:51 -0700 Subject: mprotect: prevent alteration of the PAT bits There is a defect in mprotect, which lets the user change the page cache type bits by-passing the kernel reserve_memtype and free_memtype wrappers. Fix the problem by not letting mprotect change the PAT bits. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Suresh Siddha Signed-off-by: Ingo Molnar Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-x86/pgtable.h | 16 +++++++++++++--- mm/mprotect.c | 11 ++++++++++- 2 files changed, 23 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/include/asm-x86/pgtable.h b/include/asm-x86/pgtable.h index 801b31f71452..55c3a0e3a8ce 100644 --- a/include/asm-x86/pgtable.h +++ b/include/asm-x86/pgtable.h @@ -57,7 +57,8 @@ #define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | \ _PAGE_DIRTY) -#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY) +#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_PCD | _PAGE_PWT | \ + _PAGE_ACCESSED | _PAGE_DIRTY) #define _PAGE_CACHE_MASK (_PAGE_PCD | _PAGE_PWT) #define _PAGE_CACHE_WB (0) @@ -288,12 +289,21 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot) * Chop off the NX bit (if present), and add the NX portion of * the newprot (if present): */ - val &= _PAGE_CHG_MASK & ~_PAGE_NX; - val |= pgprot_val(newprot) & __supported_pte_mask; + val &= _PAGE_CHG_MASK; + val |= pgprot_val(newprot) & (~_PAGE_CHG_MASK) & __supported_pte_mask; return __pte(val); } +/* mprotect needs to preserve PAT bits when updating vm_page_prot */ +#define pgprot_modify pgprot_modify +static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot) +{ + pgprotval_t preservebits = pgprot_val(oldprot) & _PAGE_CHG_MASK; + pgprotval_t addbits = pgprot_val(newprot); + return __pgprot(preservebits | addbits); +} + #define pte_pgprot(x) __pgprot(pte_val(x) & (0xfff | _PAGE_NX)) #define canon_pgprot(p) __pgprot(pgprot_val(p) & __supported_pte_mask) diff --git a/mm/mprotect.c b/mm/mprotect.c index 4de546899dc1..a5bf31c27375 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -26,6 +26,13 @@ #include #include +#ifndef pgprot_modify +static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot) +{ + return newprot; +} +#endif + static void change_pte_range(struct mm_struct *mm, pmd_t *pmd, unsigned long addr, unsigned long end, pgprot_t newprot, int dirty_accountable) @@ -192,7 +199,9 @@ success: * held in write mode. */ vma->vm_flags = newflags; - vma->vm_page_prot = vm_get_page_prot(newflags); + vma->vm_page_prot = pgprot_modify(vma->vm_page_prot, + vm_get_page_prot(newflags)); + if (vma_wants_writenotify(vma)) { vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED); dirty_accountable = 1; -- cgit From cec08e7a948326b01555be6311480aa08e637de2 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 30 Apr 2008 15:41:48 +1000 Subject: [POWERPC] vmemmap fixes to use smaller pages This changes vmemmap to use a different region (region 0xf) of the address space, and to configure the page size of that region dynamically at boot. The problem with the current approach of always using 16M pages is that it's not well suited to machines that have small amounts of memory such as small partitions on pseries, or PS3's. In fact, on the PS3, failure to allocate the 16M page backing vmmemmap tends to prevent hotplugging the HV's "additional" memory, thus limiting the available memory even more, from my experience down to something like 80M total, which makes it really not very useable. The logic used by my match to choose the vmemmap page size is: - If 16M pages are available and there's 1G or more RAM at boot, use that size. - Else if 64K pages are available, use that - Else use 4K pages I've tested on a POWER6 (16M pages) and on an iSeries POWER3 (4K pages) and it seems to work fine. Note that I intend to change the way we organize the kernel regions & SLBs so the actual region will change from 0xf back to something else at one point, as I simplify the SLB miss handler, but that will be for a later patch. Signed-off-by: Paul Mackerras --- arch/powerpc/mm/hash_utils_64.c | 28 ++++++++++++++++++++++++++-- arch/powerpc/mm/init_64.c | 10 ++++++---- arch/powerpc/mm/slb.c | 16 ++++++++++++++-- arch/powerpc/mm/slb_low.S | 16 +++++++++++++--- include/asm-powerpc/mmu-hash64.h | 1 + include/asm-powerpc/pgtable-ppc64.h | 10 +++++----- 6 files changed, 65 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c index 2b5a399f6fa6..0f2d239d94c4 100644 --- a/arch/powerpc/mm/hash_utils_64.c +++ b/arch/powerpc/mm/hash_utils_64.c @@ -94,6 +94,9 @@ unsigned long htab_hash_mask; int mmu_linear_psize = MMU_PAGE_4K; int mmu_virtual_psize = MMU_PAGE_4K; int mmu_vmalloc_psize = MMU_PAGE_4K; +#ifdef CONFIG_SPARSEMEM_VMEMMAP +int mmu_vmemmap_psize = MMU_PAGE_4K; +#endif int mmu_io_psize = MMU_PAGE_4K; int mmu_kernel_ssize = MMU_SEGSIZE_256M; int mmu_highuser_ssize = MMU_SEGSIZE_256M; @@ -387,11 +390,32 @@ static void __init htab_init_page_sizes(void) } #endif /* CONFIG_PPC_64K_PAGES */ +#ifdef CONFIG_SPARSEMEM_VMEMMAP + /* We try to use 16M pages for vmemmap if that is supported + * and we have at least 1G of RAM at boot + */ + if (mmu_psize_defs[MMU_PAGE_16M].shift && + lmb_phys_mem_size() >= 0x40000000) + mmu_vmemmap_psize = MMU_PAGE_16M; + else if (mmu_psize_defs[MMU_PAGE_64K].shift) + mmu_vmemmap_psize = MMU_PAGE_64K; + else + mmu_vmemmap_psize = MMU_PAGE_4K; +#endif /* CONFIG_SPARSEMEM_VMEMMAP */ + printk(KERN_DEBUG "Page orders: linear mapping = %d, " - "virtual = %d, io = %d\n", + "virtual = %d, io = %d" +#ifdef CONFIG_SPARSEMEM_VMEMMAP + ", vmemmap = %d" +#endif + "\n", mmu_psize_defs[mmu_linear_psize].shift, mmu_psize_defs[mmu_virtual_psize].shift, - mmu_psize_defs[mmu_io_psize].shift); + mmu_psize_defs[mmu_io_psize].shift +#ifdef CONFIG_SPARSEMEM_VMEMMAP + ,mmu_psize_defs[mmu_vmemmap_psize].shift +#endif + ); #ifdef CONFIG_HUGETLB_PAGE /* Init large page size. Currently, we pick 16M or 1M depending diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c index c5ac532a0161..6aa65375abf5 100644 --- a/arch/powerpc/mm/init_64.c +++ b/arch/powerpc/mm/init_64.c @@ -19,6 +19,8 @@ * */ +#undef DEBUG + #include #include #include @@ -208,12 +210,12 @@ int __meminit vmemmap_populated(unsigned long start, int page_size) } int __meminit vmemmap_populate(struct page *start_page, - unsigned long nr_pages, int node) + unsigned long nr_pages, int node) { unsigned long mode_rw; unsigned long start = (unsigned long)start_page; unsigned long end = (unsigned long)(start_page + nr_pages); - unsigned long page_size = 1 << mmu_psize_defs[mmu_linear_psize].shift; + unsigned long page_size = 1 << mmu_psize_defs[mmu_vmemmap_psize].shift; mode_rw = _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX; @@ -235,11 +237,11 @@ int __meminit vmemmap_populate(struct page *start_page, start, p, __pa(p)); mapped = htab_bolt_mapping(start, start + page_size, - __pa(p), mode_rw, mmu_linear_psize, + __pa(p), mode_rw, mmu_vmemmap_psize, mmu_kernel_ssize); BUG_ON(mapped < 0); } return 0; } -#endif +#endif /* CONFIG_SPARSEMEM_VMEMMAP */ diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c index cf8705e32d60..89497fb04280 100644 --- a/arch/powerpc/mm/slb.c +++ b/arch/powerpc/mm/slb.c @@ -28,7 +28,7 @@ #include #ifdef DEBUG -#define DBG(fmt...) udbg_printf(fmt) +#define DBG(fmt...) printk(fmt) #else #define DBG pr_debug #endif @@ -263,13 +263,19 @@ void slb_initialize(void) extern unsigned int *slb_miss_kernel_load_linear; extern unsigned int *slb_miss_kernel_load_io; extern unsigned int *slb_compare_rr_to_size; +#ifdef CONFIG_SPARSEMEM_VMEMMAP + extern unsigned int *slb_miss_kernel_load_vmemmap; + unsigned long vmemmap_llp; +#endif /* Prepare our SLB miss handler based on our page size */ linear_llp = mmu_psize_defs[mmu_linear_psize].sllp; io_llp = mmu_psize_defs[mmu_io_psize].sllp; vmalloc_llp = mmu_psize_defs[mmu_vmalloc_psize].sllp; get_paca()->vmalloc_sllp = SLB_VSID_KERNEL | vmalloc_llp; - +#ifdef CONFIG_SPARSEMEM_VMEMMAP + vmemmap_llp = mmu_psize_defs[mmu_vmemmap_psize].sllp; +#endif if (!slb_encoding_inited) { slb_encoding_inited = 1; patch_slb_encoding(slb_miss_kernel_load_linear, @@ -281,6 +287,12 @@ void slb_initialize(void) DBG("SLB: linear LLP = %04lx\n", linear_llp); DBG("SLB: io LLP = %04lx\n", io_llp); + +#ifdef CONFIG_SPARSEMEM_VMEMMAP + patch_slb_encoding(slb_miss_kernel_load_vmemmap, + SLB_VSID_KERNEL | vmemmap_llp); + DBG("SLB: vmemmap LLP = %04lx\n", vmemmap_llp); +#endif } get_paca()->stab_rr = SLB_NUM_BOLTED; diff --git a/arch/powerpc/mm/slb_low.S b/arch/powerpc/mm/slb_low.S index 657f6b37e9df..bc44dc4b5c67 100644 --- a/arch/powerpc/mm/slb_low.S +++ b/arch/powerpc/mm/slb_low.S @@ -47,8 +47,7 @@ _GLOBAL(slb_allocate_realmode) * it to VSID 0, which is reserved as a bad VSID - one which * will never have any pages in it. */ - /* Check if hitting the linear mapping of the vmalloc/ioremap - * kernel space + /* Check if hitting the linear mapping or some other kernel space */ bne cr7,1f @@ -62,7 +61,18 @@ BEGIN_FTR_SECTION END_FTR_SECTION_IFCLR(CPU_FTR_1T_SEGMENT) b slb_finish_load_1T -1: /* vmalloc/ioremap mapping encoding bits, the "li" instructions below +1: +#ifdef CONFIG_SPARSEMEM_VMEMMAP + /* Check virtual memmap region. To be patches at kernel boot */ + cmpldi cr0,r9,0xf + bne 1f +_GLOBAL(slb_miss_kernel_load_vmemmap) + li r11,0 + b 6f +1: +#endif /* CONFIG_SPARSEMEM_VMEMMAP */ + + /* vmalloc/ioremap mapping encoding bits, the "li" instructions below * will be patched by the kernel at boot */ BEGIN_FTR_SECTION diff --git a/include/asm-powerpc/mmu-hash64.h b/include/asm-powerpc/mmu-hash64.h index 0dff76776044..39c5c5f62bf5 100644 --- a/include/asm-powerpc/mmu-hash64.h +++ b/include/asm-powerpc/mmu-hash64.h @@ -177,6 +177,7 @@ extern struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT]; extern int mmu_linear_psize; extern int mmu_virtual_psize; extern int mmu_vmalloc_psize; +extern int mmu_vmemmap_psize; extern int mmu_io_psize; extern int mmu_kernel_ssize; extern int mmu_highuser_ssize; diff --git a/include/asm-powerpc/pgtable-ppc64.h b/include/asm-powerpc/pgtable-ppc64.h index 27f18695f7d6..cc6a43ba41d0 100644 --- a/include/asm-powerpc/pgtable-ppc64.h +++ b/include/asm-powerpc/pgtable-ppc64.h @@ -65,15 +65,15 @@ #define VMALLOC_REGION_ID (REGION_ID(VMALLOC_START)) #define KERNEL_REGION_ID (REGION_ID(PAGE_OFFSET)) +#define VMEMMAP_REGION_ID (0xfUL) #define USER_REGION_ID (0UL) /* - * Defines the address of the vmemap area, in the top 16th of the - * kernel region. + * Defines the address of the vmemap area, in its own region */ -#define VMEMMAP_BASE (ASM_CONST(CONFIG_KERNEL_START) + \ - (0xfUL << (REGION_SHIFT - 4))) -#define vmemmap ((struct page *)VMEMMAP_BASE) +#define VMEMMAP_BASE (VMEMMAP_REGION_ID << REGION_SHIFT) +#define vmemmap ((struct page *)VMEMMAP_BASE) + /* * Common bits in a linux-style PTE. These match the bits in the -- cgit From 9c8387afdc93f90bf0241411d44e011d8d5b76df Mon Sep 17 00:00:00 2001 From: Nate Case Date: Tue, 13 May 2008 06:14:14 +1000 Subject: [POWERPC] Fix uninitialized variable bug in copy_{to|from}_user Calls to copy_to_user() or copy_from_user() can fail when copying N bytes, where N is a constant less than 8, but not 1, 2, 4, or 8, because 'ret' is not initialized and is only set if the size is 1, 2, 4 or 8, but is tested after the switch statement for any constant size <= 8. This fixes it by initializing 'ret' to 1, causing the code to fall through to the __copy_tofrom_user call for sizes other than 1, 2, 4 or 8. Signed-off-by: Dave Scidmore Signed-off-by: Nate Case Signed-off-by: Paul Mackerras --- include/asm-powerpc/uaccess.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/asm-powerpc/uaccess.h b/include/asm-powerpc/uaccess.h index 8e798e3758bc..1a0736f8803f 100644 --- a/include/asm-powerpc/uaccess.h +++ b/include/asm-powerpc/uaccess.h @@ -380,7 +380,7 @@ static inline unsigned long __copy_from_user_inatomic(void *to, const void __user *from, unsigned long n) { if (__builtin_constant_p(n) && (n <= 8)) { - unsigned long ret; + unsigned long ret = 1; switch (n) { case 1: @@ -406,7 +406,7 @@ static inline unsigned long __copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) { if (__builtin_constant_p(n) && (n <= 8)) { - unsigned long ret; + unsigned long ret = 1; switch (n) { case 1: -- cgit From 9e491e54f0589cc26e2c096664e9d95493b1af29 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Wed, 14 May 2008 16:21:54 -0700 Subject: parisc: new termios definitions Signed-off-by: Alan Cox Cc: Kyle McMartin Cc: Grant Grundler Cc: Matthew Wilcox Signed-off-by: Andrew Morton Signed-off-by: Kyle McMartin --- include/asm-parisc/ioctls.h | 4 ++++ include/asm-parisc/termbits.h | 5 ++++- include/asm-parisc/termios.h | 6 ++++-- 3 files changed, 12 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/asm-parisc/ioctls.h b/include/asm-parisc/ioctls.h index ee84e4172c36..6747fad07a3e 100644 --- a/include/asm-parisc/ioctls.h +++ b/include/asm-parisc/ioctls.h @@ -46,6 +46,10 @@ #define TIOCSBRK 0x5427 /* BSD compatibility */ #define TIOCCBRK 0x5428 /* BSD compatibility */ #define TIOCGSID _IOR('T', 20, int) /* Return the session ID of FD */ +#define TCGETS2 _IOR('T',0x2A, struct termios2) +#define TCSETS2 _IOW('T',0x2B, struct termios2) +#define TCSETSW2 _IOW('T',0x2C, struct termios2) +#define TCSETSF2 _IOW('T',0x2D, struct termios2) #define TIOCGPTN _IOR('T',0x30, unsigned int) /* Get Pty Number (of pty-mux device) */ #define TIOCSPTLCK _IOW('T',0x31, int) /* Lock/unlock Pty */ diff --git a/include/asm-parisc/termbits.h b/include/asm-parisc/termbits.h index e847fe979684..d8bbc73b16b7 100644 --- a/include/asm-parisc/termbits.h +++ b/include/asm-parisc/termbits.h @@ -141,6 +141,7 @@ struct ktermios { #define HUPCL 0002000 #define CLOCAL 0004000 #define CBAUDEX 0010000 +#define BOTHER 0010000 #define B57600 0010001 #define B115200 0010002 #define B230400 0010003 @@ -156,10 +157,12 @@ struct ktermios { #define B3000000 0010015 #define B3500000 0010016 #define B4000000 0010017 -#define CIBAUD 002003600000 /* input baud rate (not used) */ +#define CIBAUD 002003600000 /* input baud rate */ #define CMSPAR 010000000000 /* mark or space (stick) parity */ #define CRTSCTS 020000000000 /* flow control */ +#define IBSHIFT 16 /* Shift from CBAUD to CIBAUD */ + /* c_lflag bits */ #define ISIG 0000001 diff --git a/include/asm-parisc/termios.h b/include/asm-parisc/termios.h index 5345b3420475..a2a57a4548af 100644 --- a/include/asm-parisc/termios.h +++ b/include/asm-parisc/termios.h @@ -80,8 +80,10 @@ struct termio { copy_to_user((termio)->c_cc, (termios)->c_cc, NCC); \ }) -#define user_termios_to_kernel_termios(k, u) copy_from_user(k, u, sizeof(struct termios)) -#define kernel_termios_to_user_termios(u, k) copy_to_user(u, k, sizeof(struct termios)) +#define user_termios_to_kernel_termios(k, u) copy_from_user(k, u, sizeof(struct termios2)) +#define kernel_termios_to_user_termios(u, k) copy_to_user(u, k, sizeof(struct termios2)) +#define user_termios_to_kernel_termios_1(k, u) copy_from_user(k, u, sizeof(struct termios)) +#define kernel_termios_to_user_termios_1(u, k) copy_to_user(u, k, sizeof(struct termios)) #endif /* __KERNEL__ */ -- cgit From 5cbbf16a0fab91662af8400b5ada658990932a87 Mon Sep 17 00:00:00 2001 From: Cornelia Huck Date: Thu, 15 May 2008 16:52:35 +0200 Subject: [S390] s390dbf: Use const char * for dbf name. We should use const char * for passing the name of the debug feature around since it will not be changed. Signed-off-by: Cornelia Huck Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/debug.c | 20 ++++++++++---------- include/asm-s390/debug.h | 4 ++-- 2 files changed, 12 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index dff0568e67ec..c93d1296cc0a 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -71,7 +71,7 @@ static ssize_t debug_input(struct file *file, const char __user *user_buf, size_t user_len, loff_t * offset); static int debug_open(struct inode *inode, struct file *file); static int debug_close(struct inode *inode, struct file *file); -static debug_info_t* debug_info_create(char *name, int pages_per_area, +static debug_info_t *debug_info_create(const char *name, int pages_per_area, int nr_areas, int buf_size, mode_t mode); static void debug_info_get(debug_info_t *); static void debug_info_put(debug_info_t *); @@ -234,8 +234,8 @@ fail_malloc_areas: */ static debug_info_t* -debug_info_alloc(char *name, int pages_per_area, int nr_areas, int buf_size, - int level, int mode) +debug_info_alloc(const char *name, int pages_per_area, int nr_areas, + int buf_size, int level, int mode) { debug_info_t* rc; @@ -326,8 +326,8 @@ debug_info_free(debug_info_t* db_info){ */ static debug_info_t* -debug_info_create(char *name, int pages_per_area, int nr_areas, int buf_size, - mode_t mode) +debug_info_create(const char *name, int pages_per_area, int nr_areas, + int buf_size, mode_t mode) { debug_info_t* rc; @@ -684,9 +684,9 @@ debug_close(struct inode *inode, struct file *file) * - Returns handle for debug area */ -debug_info_t *debug_register_mode(char *name, int pages_per_area, int nr_areas, - int buf_size, mode_t mode, uid_t uid, - gid_t gid) +debug_info_t *debug_register_mode(const char *name, int pages_per_area, + int nr_areas, int buf_size, mode_t mode, + uid_t uid, gid_t gid) { debug_info_t *rc = NULL; @@ -722,8 +722,8 @@ EXPORT_SYMBOL(debug_register_mode); * - returns handle for debug area */ -debug_info_t *debug_register(char *name, int pages_per_area, int nr_areas, - int buf_size) +debug_info_t *debug_register(const char *name, int pages_per_area, + int nr_areas, int buf_size) { return debug_register_mode(name, pages_per_area, nr_areas, buf_size, S_IRUSR | S_IWUSR, 0, 0); diff --git a/include/asm-s390/debug.h b/include/asm-s390/debug.h index 335baf4fc64f..9450ce6e32de 100644 --- a/include/asm-s390/debug.h +++ b/include/asm-s390/debug.h @@ -120,10 +120,10 @@ debug_entry_t* debug_exception_common(debug_info_t* id, int level, /* Debug Feature API: */ -debug_info_t* debug_register(char* name, int pages, int nr_areas, +debug_info_t *debug_register(const char *name, int pages, int nr_areas, int buf_size); -debug_info_t *debug_register_mode(char *name, int pages, int nr_areas, +debug_info_t *debug_register_mode(const char *name, int pages, int nr_areas, int buf_size, mode_t mode, uid_t uid, gid_t gid); -- cgit From f54d8a1b3fef79bb1aa2f0840dd356ce7bb180f9 Mon Sep 17 00:00:00 2001 From: "Robert P. J. Day" Date: Wed, 14 May 2008 15:14:52 -0400 Subject: parisc: Remove ioctl.h content picked up from . Now that allows overriding of the most commonly changed macro values, take advantage of that. Signed-off-by: Robert P. J. Day Signed-off-by: Kyle McMartin --- include/asm-parisc/ioctl.h | 51 +--------------------------------------------- 1 file changed, 1 insertion(+), 50 deletions(-) (limited to 'include') diff --git a/include/asm-parisc/ioctl.h b/include/asm-parisc/ioctl.h index 68338d2bda4e..ec8efa02beda 100644 --- a/include/asm-parisc/ioctl.h +++ b/include/asm-parisc/ioctl.h @@ -32,21 +32,6 @@ * NOTE: This limits the max parameter size to 16kB -1 ! */ -#define _IOC_NRBITS 8 -#define _IOC_TYPEBITS 8 -#define _IOC_SIZEBITS 14 -#define _IOC_DIRBITS 2 - -#define _IOC_NRMASK ((1 << _IOC_NRBITS)-1) -#define _IOC_TYPEMASK ((1 << _IOC_TYPEBITS)-1) -#define _IOC_SIZEMASK ((1 << _IOC_SIZEBITS)-1) -#define _IOC_DIRMASK ((1 << _IOC_DIRBITS)-1) - -#define _IOC_NRSHIFT 0 -#define _IOC_TYPESHIFT (_IOC_NRSHIFT+_IOC_NRBITS) -#define _IOC_SIZESHIFT (_IOC_TYPESHIFT+_IOC_TYPEBITS) -#define _IOC_DIRSHIFT (_IOC_SIZESHIFT+_IOC_SIZEBITS) - /* * Direction bits. */ @@ -54,40 +39,6 @@ #define _IOC_WRITE 2U #define _IOC_READ 1U -#define _IOC(dir,type,nr,size) \ - (((dir) << _IOC_DIRSHIFT) | \ - ((type) << _IOC_TYPESHIFT) | \ - ((nr) << _IOC_NRSHIFT) | \ - ((size) << _IOC_SIZESHIFT)) - -/* provoke compile error for invalid uses of size argument */ -extern unsigned int __invalid_size_argument_for_IOC; -#define _IOC_TYPECHECK(t) \ - ((sizeof(t) == sizeof(t[1]) && \ - sizeof(t) < (1 << _IOC_SIZEBITS)) ? \ - sizeof(t) : __invalid_size_argument_for_IOC) - -/* used to create numbers */ -#define _IO(type,nr) _IOC(_IOC_NONE,(type),(nr),0) -#define _IOR(type,nr,size) _IOC(_IOC_READ,(type),(nr),(_IOC_TYPECHECK(size))) -#define _IOW(type,nr,size) _IOC(_IOC_WRITE,(type),(nr),(_IOC_TYPECHECK(size))) -#define _IOWR(type,nr,size) _IOC(_IOC_READ|_IOC_WRITE,(type),(nr),(_IOC_TYPECHECK(size))) -#define _IOR_BAD(type,nr,size) _IOC(_IOC_READ,(type),(nr),sizeof(size)) -#define _IOW_BAD(type,nr,size) _IOC(_IOC_WRITE,(type),(nr),sizeof(size)) -#define _IOWR_BAD(type,nr,size) _IOC(_IOC_READ|_IOC_WRITE,(type),(nr),sizeof(size)) - -/* used to decode ioctl numbers.. */ -#define _IOC_DIR(nr) (((nr) >> _IOC_DIRSHIFT) & _IOC_DIRMASK) -#define _IOC_TYPE(nr) (((nr) >> _IOC_TYPESHIFT) & _IOC_TYPEMASK) -#define _IOC_NR(nr) (((nr) >> _IOC_NRSHIFT) & _IOC_NRMASK) -#define _IOC_SIZE(nr) (((nr) >> _IOC_SIZESHIFT) & _IOC_SIZEMASK) - -/* ...and for the drivers/sound files... */ - -#define IOC_IN (_IOC_WRITE << _IOC_DIRSHIFT) -#define IOC_OUT (_IOC_READ << _IOC_DIRSHIFT) -#define IOC_INOUT ((_IOC_WRITE|_IOC_READ) << _IOC_DIRSHIFT) -#define IOCSIZE_MASK (_IOC_SIZEMASK << _IOC_SIZESHIFT) -#define IOCSIZE_SHIFT (_IOC_SIZESHIFT) +#include #endif /* _ASM_PARISC_IOCTL_H */ -- cgit From 872f6debcae63309eb39bfc2cc9462fb83450ee0 Mon Sep 17 00:00:00 2001 From: Kyle McMartin Date: Thu, 15 May 2008 10:53:57 -0400 Subject: parisc: use conditional macro for 64-bit wide ops This work enables us to remove -traditional from $AFLAGS on parisc. Signed-off-by: Kyle McMartin --- arch/parisc/kernel/entry.S | 46 ++++++++++++---------------- arch/parisc/kernel/pacache.S | 70 +++++++++++++++++++++---------------------- include/asm-parisc/assembly.h | 6 ++-- 3 files changed, 56 insertions(+), 66 deletions(-) (limited to 'include') diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index 1a3935e61ab7..5d0837458c19 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -41,16 +41,8 @@ #include #ifdef CONFIG_64BIT -#define CMPIB cmpib,* -#define CMPB cmpb,* -#define COND(x) *x - .level 2.0w #else -#define CMPIB cmpib, -#define CMPB cmpb, -#define COND(x) x - .level 2.0 #endif @@ -958,9 +950,9 @@ intr_check_sig: * Only do signals if we are returning to user space */ LDREG PT_IASQ0(%r16), %r20 - CMPIB=,n 0,%r20,intr_restore /* backward */ + cmpib,COND(=),n 0,%r20,intr_restore /* backward */ LDREG PT_IASQ1(%r16), %r20 - CMPIB=,n 0,%r20,intr_restore /* backward */ + cmpib,COND(=),n 0,%r20,intr_restore /* backward */ copy %r0, %r25 /* long in_syscall = 0 */ #ifdef CONFIG_64BIT @@ -1014,10 +1006,10 @@ intr_do_resched: * we jump back to intr_restore. */ LDREG PT_IASQ0(%r16), %r20 - CMPIB= 0, %r20, intr_do_preempt + cmpib,COND(=) 0, %r20, intr_do_preempt nop LDREG PT_IASQ1(%r16), %r20 - CMPIB= 0, %r20, intr_do_preempt + cmpib,COND(=) 0, %r20, intr_do_preempt nop #ifdef CONFIG_64BIT @@ -1046,7 +1038,7 @@ intr_do_preempt: /* current_thread_info()->preempt_count */ mfctl %cr30, %r1 LDREG TI_PRE_COUNT(%r1), %r19 - CMPIB<> 0, %r19, intr_restore /* if preempt_count > 0 */ + cmpib,COND(<>) 0, %r19, intr_restore /* if preempt_count > 0 */ nop /* prev insn branched backwards */ /* check if we interrupted a critical path */ @@ -1065,7 +1057,7 @@ intr_do_preempt: */ intr_extint: - CMPIB=,n 0,%r16,1f + cmpib,COND(=),n 0,%r16,1f get_stack_use_cr30 b,n 2f @@ -1100,7 +1092,7 @@ ENDPROC(syscall_exit_rfi) ENTRY(intr_save) /* for os_hpmc */ mfsp %sr7,%r16 - CMPIB=,n 0,%r16,1f + cmpib,COND(=),n 0,%r16,1f get_stack_use_cr30 b 2f copy %r8,%r26 @@ -1122,7 +1114,7 @@ ENTRY(intr_save) /* for os_hpmc */ * adjust isr/ior below. */ - CMPIB=,n 6,%r26,skip_save_ior + cmpib,COND(=),n 6,%r26,skip_save_ior mfctl %cr20, %r16 /* isr */ @@ -1451,11 +1443,11 @@ nadtlb_emulate: bb,>=,n %r9,26,nadtlb_nullify /* m bit not set, just nullify */ BL get_register,%r25 extrw,u %r9,15,5,%r8 /* Get index register # */ - CMPIB=,n -1,%r1,nadtlb_fault /* have to use slow path */ + cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ copy %r1,%r24 BL get_register,%r25 extrw,u %r9,10,5,%r8 /* Get base register # */ - CMPIB=,n -1,%r1,nadtlb_fault /* have to use slow path */ + cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ BL set_register,%r25 add,l %r1,%r24,%r1 /* doesn't affect c/b bits */ @@ -1487,7 +1479,7 @@ nadtlb_probe_check: cmpb,<>,n %r16,%r17,nadtlb_fault /* Must be probe,[rw]*/ BL get_register,%r25 /* Find the target register */ extrw,u %r9,31,5,%r8 /* Get target register */ - CMPIB=,n -1,%r1,nadtlb_fault /* have to use slow path */ + cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ BL set_register,%r25 copy %r0,%r1 /* Write zero to target register */ b nadtlb_nullify /* Nullify return insn */ @@ -1571,12 +1563,12 @@ dbit_trap_20w: L3_ptep ptp,pte,t0,va,dbit_fault #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nolock_20w + cmpib,COND(=),n 0,spc,dbit_nolock_20w load32 PA(pa_dbit_lock),t0 dbit_spin_20w: LDCW 0(t0),t1 - cmpib,= 0,t1,dbit_spin_20w + cmpib,COND(=) 0,t1,dbit_spin_20w nop dbit_nolock_20w: @@ -1587,7 +1579,7 @@ dbit_nolock_20w: idtlbt pte,prot #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nounlock_20w + cmpib,COND(=),n 0,spc,dbit_nounlock_20w ldi 1,t1 stw t1,0(t0) @@ -1607,7 +1599,7 @@ dbit_trap_11: L2_ptep ptp,pte,t0,va,dbit_fault #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nolock_11 + cmpib,COND(=),n 0,spc,dbit_nolock_11 load32 PA(pa_dbit_lock),t0 dbit_spin_11: @@ -1629,7 +1621,7 @@ dbit_nolock_11: mtsp t1, %sr1 /* Restore sr1 */ #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nounlock_11 + cmpib,COND(=),n 0,spc,dbit_nounlock_11 ldi 1,t1 stw t1,0(t0) @@ -1647,7 +1639,7 @@ dbit_trap_20: L2_ptep ptp,pte,t0,va,dbit_fault #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nolock_20 + cmpib,COND(=),n 0,spc,dbit_nolock_20 load32 PA(pa_dbit_lock),t0 dbit_spin_20: @@ -1666,7 +1658,7 @@ dbit_nolock_20: idtlbt pte,prot #ifdef CONFIG_SMP - CMPIB=,n 0,spc,dbit_nounlock_20 + cmpib,COND(=),n 0,spc,dbit_nounlock_20 ldi 1,t1 stw t1,0(t0) @@ -1995,7 +1987,7 @@ ENTRY(syscall_exit) /* We can't use "CMPIB<> PER_HPUX" since "im5" field is sign extended */ ldo -PER_HPUX(%r19), %r19 - CMPIB<>,n 0,%r19,1f + cmpib,COND(<>),n 0,%r19,1f /* Save other hpux returns if personality is PER_HPUX */ STREG %r22,TASK_PT_GR22(%r1) diff --git a/arch/parisc/kernel/pacache.S b/arch/parisc/kernel/pacache.S index 7e4a33978907..e3246a5ca74f 100644 --- a/arch/parisc/kernel/pacache.S +++ b/arch/parisc/kernel/pacache.S @@ -86,7 +86,7 @@ ENTRY(flush_tlb_all_local) LDREG ITLB_OFF_COUNT(%r1), %arg2 LDREG ITLB_LOOP(%r1), %arg3 - ADDIB= -1, %arg3, fitoneloop /* Preadjust and test */ + addib,COND(=) -1, %arg3, fitoneloop /* Preadjust and test */ movb,<,n %arg3, %r31, fitdone /* If loop < 0, skip */ copy %arg0, %r28 /* Init base addr */ @@ -96,14 +96,14 @@ fitmanyloop: /* Loop if LOOP >= 2 */ copy %arg2, %r29 /* Init middle loop count */ fitmanymiddle: /* Loop if LOOP >= 2 */ - ADDIB> -1, %r31, fitmanymiddle /* Adjusted inner loop decr */ + addib,COND(>) -1, %r31, fitmanymiddle /* Adjusted inner loop decr */ pitlbe 0(%sr1, %r28) pitlbe,m %arg1(%sr1, %r28) /* Last pitlbe and addr adjust */ - ADDIB> -1, %r29, fitmanymiddle /* Middle loop decr */ + addib,COND(>) -1, %r29, fitmanymiddle /* Middle loop decr */ copy %arg3, %r31 /* Re-init inner loop count */ movb,tr %arg0, %r28, fitmanyloop /* Re-init base addr */ - ADDIB<=,n -1, %r22, fitdone /* Outer loop count decr */ + addib,COND(<=),n -1, %r22, fitdone /* Outer loop count decr */ fitoneloop: /* Loop if LOOP = 1 */ mtsp %r20, %sr1 @@ -111,10 +111,10 @@ fitoneloop: /* Loop if LOOP = 1 */ copy %arg2, %r29 /* init middle loop count */ fitonemiddle: /* Loop if LOOP = 1 */ - ADDIB> -1, %r29, fitonemiddle /* Middle loop count decr */ + addib,COND(>) -1, %r29, fitonemiddle /* Middle loop count decr */ pitlbe,m %arg1(%sr1, %r28) /* pitlbe for one loop */ - ADDIB> -1, %r22, fitoneloop /* Outer loop count decr */ + addib,COND(>) -1, %r22, fitoneloop /* Outer loop count decr */ add %r21, %r20, %r20 /* increment space */ fitdone: @@ -129,7 +129,7 @@ fitdone: LDREG DTLB_OFF_COUNT(%r1), %arg2 LDREG DTLB_LOOP(%r1), %arg3 - ADDIB= -1, %arg3, fdtoneloop /* Preadjust and test */ + addib,COND(=) -1, %arg3, fdtoneloop /* Preadjust and test */ movb,<,n %arg3, %r31, fdtdone /* If loop < 0, skip */ copy %arg0, %r28 /* Init base addr */ @@ -139,14 +139,14 @@ fdtmanyloop: /* Loop if LOOP >= 2 */ copy %arg2, %r29 /* Init middle loop count */ fdtmanymiddle: /* Loop if LOOP >= 2 */ - ADDIB> -1, %r31, fdtmanymiddle /* Adjusted inner loop decr */ + addib,COND(>) -1, %r31, fdtmanymiddle /* Adjusted inner loop decr */ pdtlbe 0(%sr1, %r28) pdtlbe,m %arg1(%sr1, %r28) /* Last pdtlbe and addr adjust */ - ADDIB> -1, %r29, fdtmanymiddle /* Middle loop decr */ + addib,COND(>) -1, %r29, fdtmanymiddle /* Middle loop decr */ copy %arg3, %r31 /* Re-init inner loop count */ movb,tr %arg0, %r28, fdtmanyloop /* Re-init base addr */ - ADDIB<=,n -1, %r22,fdtdone /* Outer loop count decr */ + addib,COND(<=),n -1, %r22,fdtdone /* Outer loop count decr */ fdtoneloop: /* Loop if LOOP = 1 */ mtsp %r20, %sr1 @@ -154,10 +154,10 @@ fdtoneloop: /* Loop if LOOP = 1 */ copy %arg2, %r29 /* init middle loop count */ fdtonemiddle: /* Loop if LOOP = 1 */ - ADDIB> -1, %r29, fdtonemiddle /* Middle loop count decr */ + addib,COND(>) -1, %r29, fdtonemiddle /* Middle loop count decr */ pdtlbe,m %arg1(%sr1, %r28) /* pdtlbe for one loop */ - ADDIB> -1, %r22, fdtoneloop /* Outer loop count decr */ + addib,COND(>) -1, %r22, fdtoneloop /* Outer loop count decr */ add %r21, %r20, %r20 /* increment space */ @@ -210,18 +210,18 @@ ENTRY(flush_instruction_cache_local) LDREG ICACHE_COUNT(%r1), %arg2 LDREG ICACHE_LOOP(%r1), %arg3 rsm PSW_SM_I, %r22 /* No mmgt ops during loop*/ - ADDIB= -1, %arg3, fioneloop /* Preadjust and test */ + addib,COND(=) -1, %arg3, fioneloop /* Preadjust and test */ movb,<,n %arg3, %r31, fisync /* If loop < 0, do sync */ fimanyloop: /* Loop if LOOP >= 2 */ - ADDIB> -1, %r31, fimanyloop /* Adjusted inner loop decr */ + addib,COND(>) -1, %r31, fimanyloop /* Adjusted inner loop decr */ fice %r0(%sr1, %arg0) fice,m %arg1(%sr1, %arg0) /* Last fice and addr adjust */ movb,tr %arg3, %r31, fimanyloop /* Re-init inner loop count */ - ADDIB<=,n -1, %arg2, fisync /* Outer loop decr */ + addib,COND(<=),n -1, %arg2, fisync /* Outer loop decr */ fioneloop: /* Loop if LOOP = 1 */ - ADDIB> -1, %arg2, fioneloop /* Outer loop count decr */ + addib,COND(>) -1, %arg2, fioneloop /* Outer loop count decr */ fice,m %arg1(%sr1, %arg0) /* Fice for one loop */ fisync: @@ -251,18 +251,18 @@ ENTRY(flush_data_cache_local) LDREG DCACHE_COUNT(%r1), %arg2 LDREG DCACHE_LOOP(%r1), %arg3 rsm PSW_SM_I, %r22 - ADDIB= -1, %arg3, fdoneloop /* Preadjust and test */ + addib,COND(=) -1, %arg3, fdoneloop /* Preadjust and test */ movb,<,n %arg3, %r31, fdsync /* If loop < 0, do sync */ fdmanyloop: /* Loop if LOOP >= 2 */ - ADDIB> -1, %r31, fdmanyloop /* Adjusted inner loop decr */ + addib,COND(>) -1, %r31, fdmanyloop /* Adjusted inner loop decr */ fdce %r0(%sr1, %arg0) fdce,m %arg1(%sr1, %arg0) /* Last fdce and addr adjust */ movb,tr %arg3, %r31, fdmanyloop /* Re-init inner loop count */ - ADDIB<=,n -1, %arg2, fdsync /* Outer loop decr */ + addib,COND(<=),n -1, %arg2, fdsync /* Outer loop decr */ fdoneloop: /* Loop if LOOP = 1 */ - ADDIB> -1, %arg2, fdoneloop /* Outer loop count decr */ + addib,COND(>) -1, %arg2, fdoneloop /* Outer loop count decr */ fdce,m %arg1(%sr1, %arg0) /* Fdce for one loop */ fdsync: @@ -343,7 +343,7 @@ ENTRY(copy_user_page_asm) * non-taken backward branch. Note that .+4 is a backwards branch. * The ldd should only get executed if the branch is taken. */ - ADDIB>,n -1, %r1, 1b /* bundle 10 */ + addib,COND(>),n -1, %r1, 1b /* bundle 10 */ ldd 0(%r25), %r19 /* start next loads */ #else @@ -392,7 +392,7 @@ ENTRY(copy_user_page_asm) stw %r21, 56(%r26) stw %r22, 60(%r26) ldo 64(%r26), %r26 - ADDIB>,n -1, %r1, 1b + addib,COND(>),n -1, %r1, 1b ldw 0(%r25), %r19 #endif bv %r0(%r2) @@ -516,7 +516,7 @@ ENTRY(copy_user_page_asm) stw %r21, 56(%r28) stw %r22, 60(%r28) ldo 64(%r28), %r28 - ADDIB> -1, %r1,1b + addib,COND(>) -1, %r1,1b ldo 64(%r29), %r29 bv %r0(%r2) @@ -575,7 +575,7 @@ ENTRY(__clear_user_page_asm) std %r0, 104(%r28) std %r0, 112(%r28) std %r0, 120(%r28) - ADDIB> -1, %r1, 1b + addib,COND(>) -1, %r1, 1b ldo 128(%r28), %r28 #else /* ! CONFIG_64BIT */ @@ -598,7 +598,7 @@ ENTRY(__clear_user_page_asm) stw %r0, 52(%r28) stw %r0, 56(%r28) stw %r0, 60(%r28) - ADDIB> -1, %r1, 1b + addib,COND(>) -1, %r1, 1b ldo 64(%r28), %r28 #endif /* CONFIG_64BIT */ @@ -641,7 +641,7 @@ ENTRY(flush_kernel_dcache_page_asm) fdc,m %r23(%r26) fdc,m %r23(%r26) fdc,m %r23(%r26) - CMPB<< %r26, %r25,1b + cmpb,COND(<<) %r26, %r25,1b fdc,m %r23(%r26) sync @@ -684,7 +684,7 @@ ENTRY(flush_user_dcache_page) fdc,m %r23(%sr3, %r26) fdc,m %r23(%sr3, %r26) fdc,m %r23(%sr3, %r26) - CMPB<< %r26, %r25,1b + cmpb,COND(<<) %r26, %r25,1b fdc,m %r23(%sr3, %r26) sync @@ -727,7 +727,7 @@ ENTRY(flush_user_icache_page) fic,m %r23(%sr3, %r26) fic,m %r23(%sr3, %r26) fic,m %r23(%sr3, %r26) - CMPB<< %r26, %r25,1b + cmpb,COND(<<) %r26, %r25,1b fic,m %r23(%sr3, %r26) sync @@ -770,7 +770,7 @@ ENTRY(purge_kernel_dcache_page) pdc,m %r23(%r26) pdc,m %r23(%r26) pdc,m %r23(%r26) - CMPB<< %r26, %r25, 1b + cmpb,COND(<<) %r26, %r25, 1b pdc,m %r23(%r26) sync @@ -834,7 +834,7 @@ ENTRY(flush_alias_page) fdc,m %r23(%r28) fdc,m %r23(%r28) fdc,m %r23(%r28) - CMPB<< %r28, %r29, 1b + cmpb,COND(<<) %r28, %r29, 1b fdc,m %r23(%r28) sync @@ -857,7 +857,7 @@ flush_user_dcache_range_asm: ldo -1(%r23), %r21 ANDCM %r26, %r21, %r26 -1: CMPB<<,n %r26, %r25, 1b +1: cmpb,COND(<<),n %r26, %r25, 1b fdc,m %r23(%sr3, %r26) sync @@ -878,7 +878,7 @@ ENTRY(flush_kernel_dcache_range_asm) ldo -1(%r23), %r21 ANDCM %r26, %r21, %r26 -1: CMPB<<,n %r26, %r25,1b +1: cmpb,COND(<<),n %r26, %r25,1b fdc,m %r23(%r26) sync @@ -900,7 +900,7 @@ ENTRY(flush_user_icache_range_asm) ldo -1(%r23), %r21 ANDCM %r26, %r21, %r26 -1: CMPB<<,n %r26, %r25,1b +1: cmpb,COND(<<),n %r26, %r25,1b fic,m %r23(%sr3, %r26) sync @@ -943,7 +943,7 @@ ENTRY(flush_kernel_icache_page) fic,m %r23(%sr4, %r26) fic,m %r23(%sr4, %r26) fic,m %r23(%sr4, %r26) - CMPB<< %r26, %r25, 1b + cmpb,COND(<<) %r26, %r25, 1b fic,m %r23(%sr4, %r26) sync @@ -964,7 +964,7 @@ ENTRY(flush_kernel_icache_range_asm) ldo -1(%r23), %r21 ANDCM %r26, %r21, %r26 -1: CMPB<<,n %r26, %r25, 1b +1: cmpb,COND(<<),n %r26, %r25, 1b fic,m %r23(%sr4, %r26) sync diff --git a/include/asm-parisc/assembly.h b/include/asm-parisc/assembly.h index 5587f0023881..ffb208840ecc 100644 --- a/include/asm-parisc/assembly.h +++ b/include/asm-parisc/assembly.h @@ -31,9 +31,8 @@ #define STREGM std,ma #define SHRREG shrd #define SHLREG shld -#define ADDIB addib,* -#define CMPB cmpb,* #define ANDCM andcm,* +#define COND(x) * ## x #define RP_OFFSET 16 #define FRAME_SIZE 128 #define CALLEE_REG_FRAME_SIZE 144 @@ -46,9 +45,8 @@ #define STREGM stwm #define SHRREG shr #define SHLREG shlw -#define ADDIB addib, -#define CMPB cmpb, #define ANDCM andcm +#define COND(x) x #define RP_OFFSET 20 #define FRAME_SIZE 64 #define CALLEE_REG_FRAME_SIZE 128 -- cgit From a442ac512f36981182e66a427ad05f449ff6593b Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 15 May 2008 17:50:37 -0700 Subject: Clean up 'print_fn_descriptor_symbol()' types Everybody wants to pass it a function pointer, and in fact, that is what you _must_ pass it for it to make sense (since it knows that ia64 and ppc64 use descriptors for function pointers and fetches the actual address from there). So don't make the argument be a 'unsigned long' and force everybody to add a cast. Signed-off-by: Linus Torvalds --- drivers/base/power/main.c | 2 +- drivers/pci/quirks.c | 3 +-- drivers/pnp/quirks.c | 3 +-- include/linux/kallsyms.h | 24 ++++++++++++++---------- init/main.c | 9 +++------ 5 files changed, 20 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 7b76fd3b93a4..45cc3d9eacb8 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -418,7 +418,7 @@ void __suspend_report_result(const char *function, void *fn, int ret) { if (ret) { printk(KERN_ERR "%s(): ", function); - print_fn_descriptor_symbol("%s() returns ", (unsigned long)fn); + print_fn_descriptor_symbol("%s returns ", fn); printk("%d\n", ret); } } diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f2d9c770f51a..dabb563f51d9 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -1503,8 +1503,7 @@ static void pci_do_fixups(struct pci_dev *dev, struct pci_fixup *f, struct pci_f (f->device == dev->device || f->device == (u16) PCI_ANY_ID)) { #ifdef DEBUG dev_dbg(&dev->dev, "calling "); - print_fn_descriptor_symbol("%s()\n", - (unsigned long) f->hook); + print_fn_descriptor_symbol("%s\n", f->hook); #endif f->hook(dev); } diff --git a/drivers/pnp/quirks.c b/drivers/pnp/quirks.c index ffdb12a59c40..e2b7de4cb05e 100644 --- a/drivers/pnp/quirks.c +++ b/drivers/pnp/quirks.c @@ -331,8 +331,7 @@ void pnp_fixup_device(struct pnp_dev *dev) continue; #ifdef DEBUG dev_dbg(&dev->dev, "%s: calling ", f->id); - print_fn_descriptor_symbol("%s\n", - (unsigned long) f->quirk_function); + print_fn_descriptor_symbol("%s\n", f->quirk_function); #endif f->quirk_function(dev); } diff --git a/include/linux/kallsyms.h b/include/linux/kallsyms.h index 82de2fb62cb7..00c1801099fa 100644 --- a/include/linux/kallsyms.h +++ b/include/linux/kallsyms.h @@ -83,16 +83,6 @@ __attribute__((format(printf,1,2))); static inline void __check_printsym_format(const char *fmt, ...) { } -/* ia64 and ppc64 use function descriptors, which contain the real address */ -#if defined(CONFIG_IA64) || defined(CONFIG_PPC64) -#define print_fn_descriptor_symbol(fmt, addr) \ -do { \ - unsigned long *__faddr = (unsigned long*) addr; \ - print_symbol(fmt, __faddr[0]); \ -} while (0) -#else -#define print_fn_descriptor_symbol(fmt, addr) print_symbol(fmt, addr) -#endif static inline void print_symbol(const char *fmt, unsigned long addr) { @@ -101,6 +91,20 @@ static inline void print_symbol(const char *fmt, unsigned long addr) __builtin_extract_return_addr((void *)addr)); } +/* + * Pretty-print a function pointer. + * + * ia64 and ppc64 function pointers are really function descriptors, + * which contain a pointer the real address. + */ +static inline void print_fn_descriptor_symbol(const char *fmt, void *addr) +{ +#if defined(CONFIG_IA64) || defined(CONFIG_PPC64) + addr = *(void **)addr; +#endif + print_symbol(fmt, (unsigned long)addr); +} + #ifndef CONFIG_64BIT #define print_ip_sym(ip) \ do { \ diff --git a/init/main.c b/init/main.c index f406fefa626c..c62215146a80 100644 --- a/init/main.c +++ b/init/main.c @@ -706,8 +706,7 @@ static void __init do_initcalls(void) int result; if (initcall_debug) { - print_fn_descriptor_symbol("calling %s()\n", - (unsigned long) *call); + print_fn_descriptor_symbol("calling %s\n", *call); t0 = ktime_get(); } @@ -717,8 +716,7 @@ static void __init do_initcalls(void) t1 = ktime_get(); delta = ktime_sub(t1, t0); - print_fn_descriptor_symbol("initcall %s()", - (unsigned long) *call); + print_fn_descriptor_symbol("initcall %s", *call); printk(" returned %d after %Ld msecs\n", result, (unsigned long long) delta.tv64 >> 20); } @@ -737,8 +735,7 @@ static void __init do_initcalls(void) local_irq_enable(); } if (msgbuf[0]) { - print_fn_descriptor_symbol(KERN_WARNING "initcall %s()", - (unsigned long) *call); + print_fn_descriptor_symbol(KERN_WARNING "initcall %s", *call); printk(" returned with %s\n", msgbuf); } } -- cgit