From e93ad19d05648397ef3bcb838d26aec06c245dc0 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 19 Jan 2016 12:18:41 -0500 Subject: cpuset: make mm migration asynchronous If "cpuset.memory_migrate" is set, when a process is moved from one cpuset to another with a different memory node mask, pages in used by the process are migrated to the new set of nodes. This was performed synchronously in the ->attach() callback, which is synchronized against process management. Recently, the synchronization was changed from per-process rwsem to global percpu rwsem for simplicity and optimization. Combined with the synchronous mm migration, this led to deadlocks because mm migration could schedule a work item which may in turn try to create a new worker blocking on the process management lock held from cgroup process migration path. This heavy an operation shouldn't be performed synchronously from that deep inside cgroup migration in the first place. This patch punts the actual migration to an ordered workqueue and updates cgroup process migration and cpuset config update paths to flush the workqueue after all locks are released. This way, the operations still seem synchronous to userland without entangling mm migration with process management synchronization. CPU hotplug can also invoke mm migration but there's no reason for it to wait for mm migrations and thus doesn't synchronize against their completions. Signed-off-by: Tejun Heo Reported-and-tested-by: Christian Borntraeger Cc: stable@vger.kernel.org # v4.4+ --- include/linux/cpuset.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include/linux') diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h index 85a868ccb493..fea160ee5803 100644 --- a/include/linux/cpuset.h +++ b/include/linux/cpuset.h @@ -137,6 +137,8 @@ static inline void set_mems_allowed(nodemask_t nodemask) task_unlock(current); } +extern void cpuset_post_attach_flush(void); + #else /* !CONFIG_CPUSETS */ static inline bool cpusets_enabled(void) { return false; } @@ -243,6 +245,10 @@ static inline bool read_mems_allowed_retry(unsigned int seq) return false; } +static inline void cpuset_post_attach_flush(void) +{ +} + #endif /* !CONFIG_CPUSETS */ #endif /* _LINUX_CPUSET_H */ -- cgit From aa226ff4a1ce79f229c6b7a4c0a14e17fececd01 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 21 Jan 2016 15:31:11 -0500 Subject: cgroup: make sure a parent css isn't offlined before its children There are three subsystem callbacks in css shutdown path - css_offline(), css_released() and css_free(). Except for css_released(), cgroup core didn't guarantee the order of invocation. css_offline() or css_free() could be called on a parent css before its children. This behavior is unexpected and led to bugs in cpu and memory controller. This patch updates offline path so that a parent css is never offlined before its children. Each css keeps online_cnt which reaches zero iff itself and all its children are offline and offline_css() is invoked only after online_cnt reaches zero. This fixes the memory controller bug and allows the fix for cpu controller. Signed-off-by: Tejun Heo Reported-and-tested-by: Christian Borntraeger Reported-by: Brian Christiansen Link: http://lkml.kernel.org/g/5698A023.9070703@de.ibm.com Link: http://lkml.kernel.org/g/CAKB58ikDkzc8REt31WBkD99+hxNzjK4+FBmhkgS+NVrC9vjMSg@mail.gmail.com Cc: Heiko Carstens Cc: Peter Zijlstra Cc: stable@vger.kernel.org --- include/linux/cgroup-defs.h | 6 ++++++ kernel/cgroup.c | 22 +++++++++++++++++----- 2 files changed, 23 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h index 7f540f7f588d..789471dba6fb 100644 --- a/include/linux/cgroup-defs.h +++ b/include/linux/cgroup-defs.h @@ -127,6 +127,12 @@ struct cgroup_subsys_state { */ u64 serial_nr; + /* + * Incremented by online self and children. Used to guarantee that + * parents are not offlined before their children. + */ + atomic_t online_cnt; + /* percpu_ref killing and RCU release */ struct rcu_head rcu_head; struct work_struct destroy_work; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 88abd4d076d8..d01587793865 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -4760,6 +4760,7 @@ static void init_and_link_css(struct cgroup_subsys_state *css, INIT_LIST_HEAD(&css->sibling); INIT_LIST_HEAD(&css->children); css->serial_nr = css_serial_nr_next++; + atomic_set(&css->online_cnt, 0); if (cgroup_parent(cgrp)) { css->parent = cgroup_css(cgroup_parent(cgrp), ss); @@ -4782,6 +4783,10 @@ static int online_css(struct cgroup_subsys_state *css) if (!ret) { css->flags |= CSS_ONLINE; rcu_assign_pointer(css->cgroup->subsys[ss->id], css); + + atomic_inc(&css->online_cnt); + if (css->parent) + atomic_inc(&css->parent->online_cnt); } return ret; } @@ -5019,10 +5024,15 @@ static void css_killed_work_fn(struct work_struct *work) container_of(work, struct cgroup_subsys_state, destroy_work); mutex_lock(&cgroup_mutex); - offline_css(css); - mutex_unlock(&cgroup_mutex); - css_put(css); + do { + offline_css(css); + css_put(css); + /* @css can't go away while we're holding cgroup_mutex */ + css = css->parent; + } while (css && atomic_dec_and_test(&css->online_cnt)); + + mutex_unlock(&cgroup_mutex); } /* css kill confirmation processing requires process context, bounce */ @@ -5031,8 +5041,10 @@ static void css_killed_ref_fn(struct percpu_ref *ref) struct cgroup_subsys_state *css = container_of(ref, struct cgroup_subsys_state, refcnt); - INIT_WORK(&css->destroy_work, css_killed_work_fn); - queue_work(cgroup_destroy_wq, &css->destroy_work); + if (atomic_dec_and_test(&css->online_cnt)) { + INIT_WORK(&css->destroy_work, css_killed_work_fn); + queue_work(cgroup_destroy_wq, &css->destroy_work); + } } /** -- cgit From fb3296335500aaff61333df8eabbccf28761c79d Mon Sep 17 00:00:00 2001 From: Danesh Petigara Date: Mon, 11 Jan 2016 13:22:26 -0800 Subject: drivers: ata: wake port before DMA stop for ALPM The AHCI driver code stops and starts port DMA engines at will without considering the power state of the particular port. The AHCI specification isn't very clear on how to handle this scenario, leaving implementation open to interpretation. Broadcom's STB SATA host controller is unable to handle port DMA controller restarts when the port in question is in low power mode. When a port enters partial or slumber mode, its PHY is powered down. When a controller restart is requested, the controller's internal state machine expects the PHY to be brought back up by software which never happens in this case, resulting in failures. To avoid this situation, logic is added to manually wake up the port just before its DMA engine is stopped, if the port happens to be in a low power state. HBA initiated power management ensures that the port eventually returns to its configured low power state, when the link is idle (as per the conditions listed in the spec). A new host flag is also added to ensure this logic is only exercised for hosts with the above limitation. tj: Formatting changes. Signed-off-by: Danesh Petigara Reviewed-by: Markus Mayer Signed-off-by: Tejun Heo --- drivers/ata/ahci.h | 1 + drivers/ata/ahci_brcmstb.c | 1 + drivers/ata/libahci.c | 23 ++++++++++++++++++++++- include/linux/libata.h | 1 + 4 files changed, 25 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h index a4faa438889c..a44c75d4c284 100644 --- a/drivers/ata/ahci.h +++ b/drivers/ata/ahci.h @@ -250,6 +250,7 @@ enum { AHCI_HFLAG_MULTI_MSI = 0, AHCI_HFLAG_MULTI_MSIX = 0, #endif + AHCI_HFLAG_WAKE_BEFORE_STOP = (1 << 22), /* wake before DMA stop */ /* ap->flags bits */ diff --git a/drivers/ata/ahci_brcmstb.c b/drivers/ata/ahci_brcmstb.c index b36cae2fd04b..e87bcec0fd7c 100644 --- a/drivers/ata/ahci_brcmstb.c +++ b/drivers/ata/ahci_brcmstb.c @@ -317,6 +317,7 @@ static int brcm_ahci_probe(struct platform_device *pdev) if (IS_ERR(hpriv)) return PTR_ERR(hpriv); hpriv->plat_data = priv; + hpriv->flags = AHCI_HFLAG_WAKE_BEFORE_STOP; brcm_sata_alpm_init(hpriv); diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c index d61740e78d6d..284a176076f5 100644 --- a/drivers/ata/libahci.c +++ b/drivers/ata/libahci.c @@ -593,8 +593,22 @@ EXPORT_SYMBOL_GPL(ahci_start_engine); int ahci_stop_engine(struct ata_port *ap) { void __iomem *port_mmio = ahci_port_base(ap); + struct ahci_host_priv *hpriv = ap->host->private_data; u32 tmp; + /* + * On some controllers, stopping a port's DMA engine while the port + * is in ALPM state (partial or slumber) results in failures on + * subsequent DMA engine starts. For those controllers, put the + * port back in active state before stopping its DMA engine. + */ + if ((hpriv->flags & AHCI_HFLAG_WAKE_BEFORE_STOP) && + (ap->link.lpm_policy > ATA_LPM_MAX_POWER) && + ahci_set_lpm(&ap->link, ATA_LPM_MAX_POWER, ATA_LPM_WAKE_ONLY)) { + dev_err(ap->host->dev, "Failed to wake up port before engine stop\n"); + return -EIO; + } + tmp = readl(port_mmio + PORT_CMD); /* check if the HBA is idle */ @@ -689,6 +703,9 @@ static int ahci_set_lpm(struct ata_link *link, enum ata_lpm_policy policy, void __iomem *port_mmio = ahci_port_base(ap); if (policy != ATA_LPM_MAX_POWER) { + /* wakeup flag only applies to the max power policy */ + hints &= ~ATA_LPM_WAKE_ONLY; + /* * Disable interrupts on Phy Ready. This keeps us from * getting woken up due to spurious phy ready @@ -704,7 +721,8 @@ static int ahci_set_lpm(struct ata_link *link, enum ata_lpm_policy policy, u32 cmd = readl(port_mmio + PORT_CMD); if (policy == ATA_LPM_MAX_POWER || !(hints & ATA_LPM_HIPM)) { - cmd &= ~(PORT_CMD_ASP | PORT_CMD_ALPE); + if (!(hints & ATA_LPM_WAKE_ONLY)) + cmd &= ~(PORT_CMD_ASP | PORT_CMD_ALPE); cmd |= PORT_CMD_ICC_ACTIVE; writel(cmd, port_mmio + PORT_CMD); @@ -712,6 +730,9 @@ static int ahci_set_lpm(struct ata_link *link, enum ata_lpm_policy policy, /* wait 10ms to be sure we've come out of LPM state */ ata_msleep(ap, 10); + + if (hints & ATA_LPM_WAKE_ONLY) + return 0; } else { cmd |= PORT_CMD_ALPE; if (policy == ATA_LPM_MIN_POWER) diff --git a/include/linux/libata.h b/include/linux/libata.h index 851821bfd553..bec2abbd7ab2 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -526,6 +526,7 @@ enum ata_lpm_policy { enum ata_lpm_hints { ATA_LPM_EMPTY = (1 << 0), /* port empty/probing */ ATA_LPM_HIPM = (1 << 1), /* may use HIPM */ + ATA_LPM_WAKE_ONLY = (1 << 2), /* only wake up link */ }; /* forward declarations */ -- cgit From 23d11a58a9a60dcb52c8fc6494efce908b24c295 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 29 Jan 2016 05:59:46 -0500 Subject: workqueue: skip flush dependency checks for legacy workqueues fca839c00a12 ("workqueue: warn if memory reclaim tries to flush !WQ_MEM_RECLAIM workqueue") implemented flush dependency warning which triggers if a PF_MEMALLOC task or WQ_MEM_RECLAIM workqueue tries to flush a !WQ_MEM_RECLAIM workquee. This assumes that workqueues marked with WQ_MEM_RECLAIM sit in memory reclaim path and making it depend on something which may need more memory to make forward progress can lead to deadlocks. Unfortunately, workqueues created with the legacy create*_workqueue() interface always have WQ_MEM_RECLAIM regardless of whether they are depended upon memory reclaim or not. These spurious WQ_MEM_RECLAIM markings cause spurious triggering of the flush dependency checks. WARNING: CPU: 0 PID: 6 at kernel/workqueue.c:2361 check_flush_dependency+0x138/0x144() workqueue: WQ_MEM_RECLAIM deferwq:deferred_probe_work_func is flushing !WQ_MEM_RECLAIM events:lru_add_drain_per_cpu ... Workqueue: deferwq deferred_probe_work_func [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [] (show_stack) from [] (dump_stack+0x94/0xd4) [] (dump_stack) from [] (warn_slowpath_common+0x80/0xb0) [] (warn_slowpath_common) from [] (warn_slowpath_fmt+0x30/0x40) [] (warn_slowpath_fmt) from [] (check_flush_dependency+0x138/0x144) [] (check_flush_dependency) from [] (flush_work+0x50/0x15c) [] (flush_work) from [] (lru_add_drain_all+0x130/0x180) [] (lru_add_drain_all) from [] (migrate_prep+0x8/0x10) [] (migrate_prep) from [] (alloc_contig_range+0xd8/0x338) [] (alloc_contig_range) from [] (cma_alloc+0xe0/0x1ac) [] (cma_alloc) from [] (__alloc_from_contiguous+0x38/0xd8) [] (__alloc_from_contiguous) from [] (__dma_alloc+0x240/0x278) [] (__dma_alloc) from [] (arm_dma_alloc+0x54/0x5c) [] (arm_dma_alloc) from [] (dmam_alloc_coherent+0xc0/0xec) [] (dmam_alloc_coherent) from [] (ahci_port_start+0x150/0x1dc) [] (ahci_port_start) from [] (ata_host_start.part.3+0xc8/0x1c8) [] (ata_host_start.part.3) from [] (ata_host_activate+0x50/0x148) [] (ata_host_activate) from [] (ahci_host_activate+0x44/0x114) [] (ahci_host_activate) from [] (ahci_platform_init_host+0x1d8/0x3c8) [] (ahci_platform_init_host) from [] (tegra_ahci_probe+0x448/0x4e8) [] (tegra_ahci_probe) from [] (platform_drv_probe+0x50/0xac) [] (platform_drv_probe) from [] (driver_probe_device+0x214/0x2c0) [] (driver_probe_device) from [] (bus_for_each_drv+0x60/0x94) [] (bus_for_each_drv) from [] (__device_attach+0xb0/0x114) [] (__device_attach) from [] (bus_probe_device+0x84/0x8c) [] (bus_probe_device) from [] (deferred_probe_work_func+0x68/0x98) [] (deferred_probe_work_func) from [] (process_one_work+0x120/0x3f8) [] (process_one_work) from [] (worker_thread+0x38/0x55c) [] (worker_thread) from [] (kthread+0xdc/0xf4) [] (kthread) from [] (ret_from_fork+0x14/0x3c) Fix it by marking workqueues created via create*_workqueue() with __WQ_LEGACY and disabling flush dependency checks on them. Signed-off-by: Tejun Heo Reported-and-tested-by: Thierry Reding Link: http://lkml.kernel.org/g/20160126173843.GA11115@ulmo.nvidia.com Fixes: fca839c00a12 ("workqueue: warn if memory reclaim tries to flush !WQ_MEM_RECLAIM workqueue") --- include/linux/workqueue.h | 9 +++++---- kernel/workqueue.c | 3 ++- 2 files changed, 7 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index 0e32bc71245e..ca73c503b92a 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -311,6 +311,7 @@ enum { __WQ_DRAINING = 1 << 16, /* internal: workqueue is draining */ __WQ_ORDERED = 1 << 17, /* internal: workqueue is ordered */ + __WQ_LEGACY = 1 << 18, /* internal: create*_workqueue() */ WQ_MAX_ACTIVE = 512, /* I like 512, better ideas? */ WQ_MAX_UNBOUND_PER_CPU = 4, /* 4 * #cpus for unbound wq */ @@ -411,12 +412,12 @@ __alloc_workqueue_key(const char *fmt, unsigned int flags, int max_active, alloc_workqueue(fmt, WQ_UNBOUND | __WQ_ORDERED | (flags), 1, ##args) #define create_workqueue(name) \ - alloc_workqueue("%s", WQ_MEM_RECLAIM, 1, (name)) + alloc_workqueue("%s", __WQ_LEGACY | WQ_MEM_RECLAIM, 1, (name)) #define create_freezable_workqueue(name) \ - alloc_workqueue("%s", WQ_FREEZABLE | WQ_UNBOUND | WQ_MEM_RECLAIM, \ - 1, (name)) + alloc_workqueue("%s", __WQ_LEGACY | WQ_FREEZABLE | WQ_UNBOUND | \ + WQ_MEM_RECLAIM, 1, (name)) #define create_singlethread_workqueue(name) \ - alloc_ordered_workqueue("%s", WQ_MEM_RECLAIM, name) + alloc_ordered_workqueue("%s", __WQ_LEGACY | WQ_MEM_RECLAIM, name) extern void destroy_workqueue(struct workqueue_struct *wq); diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 61a0264e28f9..dc7faad63646 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2355,7 +2355,8 @@ static void check_flush_dependency(struct workqueue_struct *target_wq, WARN_ONCE(current->flags & PF_MEMALLOC, "workqueue: PF_MEMALLOC task %d(%s) is flushing !WQ_MEM_RECLAIM %s:%pf", current->pid, current->comm, target_wq->name, target_func); - WARN_ONCE(worker && (worker->current_pwq->wq->flags & WQ_MEM_RECLAIM), + WARN_ONCE(worker && ((worker->current_pwq->wq->flags & + (WQ_MEM_RECLAIM | __WQ_LEGACY)) == WQ_MEM_RECLAIM), "workqueue: WQ_MEM_RECLAIM %s:%pf is flushing !WQ_MEM_RECLAIM %s:%pf", worker->current_pwq->wq->name, worker->current_func, target_wq->name, target_func); -- cgit From 8244062ef1e54502ef55f54cced659913f244c3e Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Wed, 3 Feb 2016 16:55:26 +1030 Subject: modules: fix longstanding /proc/kallsyms vs module insertion race. For CONFIG_KALLSYMS, we keep two symbol tables and two string tables. There's one full copy, marked SHF_ALLOC and laid out at the end of the module's init section. There's also a cut-down version that only contains core symbols and strings, and lives in the module's core section. After module init (and before we free the module memory), we switch the mod->symtab, mod->num_symtab and mod->strtab to point to the core versions. We do this under the module_mutex. However, kallsyms doesn't take the module_mutex: it uses preempt_disable() and rcu tricks to walk through the modules, because it's used in the oops path. It's also used in /proc/kallsyms. There's nothing atomic about the change of these variables, so we can get the old (larger!) num_symtab and the new symtab pointer; in fact this is what I saw when trying to reproduce. By grouping these variables together, we can use a carefully-dereferenced pointer to ensure we always get one or the other (the free of the module init section is already done in an RCU callback, so that's safe). We allocate the init one at the end of the module init section, and keep the core one inside the struct module itself (it could also have been allocated at the end of the module core, but that's probably overkill). Reported-by: Weilong Chen Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=111541 Cc: stable@kernel.org Signed-off-by: Rusty Russell --- include/linux/module.h | 19 +++++---- kernel/module.c | 112 ++++++++++++++++++++++++++++++------------------- 2 files changed, 79 insertions(+), 52 deletions(-) (limited to 'include/linux') diff --git a/include/linux/module.h b/include/linux/module.h index 4560d8f1545d..2bb0c3085706 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -324,6 +324,12 @@ struct module_layout { #define __module_layout_align #endif +struct mod_kallsyms { + Elf_Sym *symtab; + unsigned int num_symtab; + char *strtab; +}; + struct module { enum module_state state; @@ -405,15 +411,10 @@ struct module { #endif #ifdef CONFIG_KALLSYMS - /* - * We keep the symbol and string tables for kallsyms. - * The core_* fields below are temporary, loader-only (they - * could really be discarded after module init). - */ - Elf_Sym *symtab, *core_symtab; - unsigned int num_symtab, core_num_syms; - char *strtab, *core_strtab; - + /* Protected by RCU and/or module_mutex: use rcu_dereference() */ + struct mod_kallsyms *kallsyms; + struct mod_kallsyms core_kallsyms; + /* Section attributes */ struct module_sect_attrs *sect_attrs; diff --git a/kernel/module.c b/kernel/module.c index 1e79d8157712..9537da37ce87 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -303,6 +303,9 @@ struct load_info { struct _ddebug *debug; unsigned int num_debug; bool sig_ok; +#ifdef CONFIG_KALLSYMS + unsigned long mod_kallsyms_init_off; +#endif struct { unsigned int sym, str, mod, vers, info, pcpu; } index; @@ -2480,10 +2483,21 @@ static void layout_symtab(struct module *mod, struct load_info *info) strsect->sh_flags |= SHF_ALLOC; strsect->sh_entsize = get_offset(mod, &mod->init_layout.size, strsect, info->index.str) | INIT_OFFSET_MASK; - mod->init_layout.size = debug_align(mod->init_layout.size); pr_debug("\t%s\n", info->secstrings + strsect->sh_name); + + /* We'll tack temporary mod_kallsyms on the end. */ + mod->init_layout.size = ALIGN(mod->init_layout.size, + __alignof__(struct mod_kallsyms)); + info->mod_kallsyms_init_off = mod->init_layout.size; + mod->init_layout.size += sizeof(struct mod_kallsyms); + mod->init_layout.size = debug_align(mod->init_layout.size); } +/* + * We use the full symtab and strtab which layout_symtab arranged to + * be appended to the init section. Later we switch to the cut-down + * core-only ones. + */ static void add_kallsyms(struct module *mod, const struct load_info *info) { unsigned int i, ndst; @@ -2492,29 +2506,34 @@ static void add_kallsyms(struct module *mod, const struct load_info *info) char *s; Elf_Shdr *symsec = &info->sechdrs[info->index.sym]; - mod->symtab = (void *)symsec->sh_addr; - mod->num_symtab = symsec->sh_size / sizeof(Elf_Sym); + /* Set up to point into init section. */ + mod->kallsyms = mod->init_layout.base + info->mod_kallsyms_init_off; + + mod->kallsyms->symtab = (void *)symsec->sh_addr; + mod->kallsyms->num_symtab = symsec->sh_size / sizeof(Elf_Sym); /* Make sure we get permanent strtab: don't use info->strtab. */ - mod->strtab = (void *)info->sechdrs[info->index.str].sh_addr; + mod->kallsyms->strtab = (void *)info->sechdrs[info->index.str].sh_addr; /* Set types up while we still have access to sections. */ - for (i = 0; i < mod->num_symtab; i++) - mod->symtab[i].st_info = elf_type(&mod->symtab[i], info); - - mod->core_symtab = dst = mod->core_layout.base + info->symoffs; - mod->core_strtab = s = mod->core_layout.base + info->stroffs; - src = mod->symtab; - for (ndst = i = 0; i < mod->num_symtab; i++) { + for (i = 0; i < mod->kallsyms->num_symtab; i++) + mod->kallsyms->symtab[i].st_info + = elf_type(&mod->kallsyms->symtab[i], info); + + /* Now populate the cut down core kallsyms for after init. */ + mod->core_kallsyms.symtab = dst = mod->core_layout.base + info->symoffs; + mod->core_kallsyms.strtab = s = mod->core_layout.base + info->stroffs; + src = mod->kallsyms->symtab; + for (ndst = i = 0; i < mod->kallsyms->num_symtab; i++) { if (i == 0 || is_core_symbol(src+i, info->sechdrs, info->hdr->e_shnum, info->index.pcpu)) { dst[ndst] = src[i]; - dst[ndst++].st_name = s - mod->core_strtab; - s += strlcpy(s, &mod->strtab[src[i].st_name], + dst[ndst++].st_name = s - mod->core_kallsyms.strtab; + s += strlcpy(s, &mod->kallsyms->strtab[src[i].st_name], KSYM_NAME_LEN) + 1; } } - mod->core_num_syms = ndst; + mod->core_kallsyms.num_symtab = ndst; } #else static inline void layout_symtab(struct module *mod, struct load_info *info) @@ -3263,9 +3282,8 @@ static noinline int do_init_module(struct module *mod) module_put(mod); trim_init_extable(mod); #ifdef CONFIG_KALLSYMS - mod->num_symtab = mod->core_num_syms; - mod->symtab = mod->core_symtab; - mod->strtab = mod->core_strtab; + /* Switch to core kallsyms now init is done: kallsyms may be walking! */ + rcu_assign_pointer(mod->kallsyms, &mod->core_kallsyms); #endif mod_tree_remove_init(mod); disable_ro_nx(&mod->init_layout); @@ -3627,9 +3645,9 @@ static inline int is_arm_mapping_symbol(const char *str) && (str[2] == '\0' || str[2] == '.'); } -static const char *symname(struct module *mod, unsigned int symnum) +static const char *symname(struct mod_kallsyms *kallsyms, unsigned int symnum) { - return mod->strtab + mod->symtab[symnum].st_name; + return kallsyms->strtab + kallsyms->symtab[symnum].st_name; } static const char *get_ksymbol(struct module *mod, @@ -3639,6 +3657,7 @@ static const char *get_ksymbol(struct module *mod, { unsigned int i, best = 0; unsigned long nextval; + struct mod_kallsyms *kallsyms = rcu_dereference_sched(mod->kallsyms); /* At worse, next value is at end of module */ if (within_module_init(addr, mod)) @@ -3648,32 +3667,32 @@ static const char *get_ksymbol(struct module *mod, /* Scan for closest preceding symbol, and next symbol. (ELF starts real symbols at 1). */ - for (i = 1; i < mod->num_symtab; i++) { - if (mod->symtab[i].st_shndx == SHN_UNDEF) + for (i = 1; i < kallsyms->num_symtab; i++) { + if (kallsyms->symtab[i].st_shndx == SHN_UNDEF) continue; /* We ignore unnamed symbols: they're uninformative * and inserted at a whim. */ - if (*symname(mod, i) == '\0' - || is_arm_mapping_symbol(symname(mod, i))) + if (*symname(kallsyms, i) == '\0' + || is_arm_mapping_symbol(symname(kallsyms, i))) continue; - if (mod->symtab[i].st_value <= addr - && mod->symtab[i].st_value > mod->symtab[best].st_value) + if (kallsyms->symtab[i].st_value <= addr + && kallsyms->symtab[i].st_value > kallsyms->symtab[best].st_value) best = i; - if (mod->symtab[i].st_value > addr - && mod->symtab[i].st_value < nextval) - nextval = mod->symtab[i].st_value; + if (kallsyms->symtab[i].st_value > addr + && kallsyms->symtab[i].st_value < nextval) + nextval = kallsyms->symtab[i].st_value; } if (!best) return NULL; if (size) - *size = nextval - mod->symtab[best].st_value; + *size = nextval - kallsyms->symtab[best].st_value; if (offset) - *offset = addr - mod->symtab[best].st_value; - return symname(mod, best); + *offset = addr - kallsyms->symtab[best].st_value; + return symname(kallsyms, best); } /* For kallsyms to ask for address resolution. NULL means not found. Careful @@ -3763,18 +3782,21 @@ int module_get_kallsym(unsigned int symnum, unsigned long *value, char *type, preempt_disable(); list_for_each_entry_rcu(mod, &modules, list) { + struct mod_kallsyms *kallsyms; + if (mod->state == MODULE_STATE_UNFORMED) continue; - if (symnum < mod->num_symtab) { - *value = mod->symtab[symnum].st_value; - *type = mod->symtab[symnum].st_info; - strlcpy(name, symname(mod, symnum), KSYM_NAME_LEN); + kallsyms = rcu_dereference_sched(mod->kallsyms); + if (symnum < kallsyms->num_symtab) { + *value = kallsyms->symtab[symnum].st_value; + *type = kallsyms->symtab[symnum].st_info; + strlcpy(name, symname(kallsyms, symnum), KSYM_NAME_LEN); strlcpy(module_name, mod->name, MODULE_NAME_LEN); *exported = is_exported(name, *value, mod); preempt_enable(); return 0; } - symnum -= mod->num_symtab; + symnum -= kallsyms->num_symtab; } preempt_enable(); return -ERANGE; @@ -3783,11 +3805,12 @@ int module_get_kallsym(unsigned int symnum, unsigned long *value, char *type, static unsigned long mod_find_symname(struct module *mod, const char *name) { unsigned int i; + struct mod_kallsyms *kallsyms = rcu_dereference_sched(mod->kallsyms); - for (i = 0; i < mod->num_symtab; i++) - if (strcmp(name, symname(mod, i)) == 0 && - mod->symtab[i].st_info != 'U') - return mod->symtab[i].st_value; + for (i = 0; i < kallsyms->num_symtab; i++) + if (strcmp(name, symname(kallsyms, i)) == 0 && + kallsyms->symtab[i].st_info != 'U') + return kallsyms->symtab[i].st_value; return 0; } @@ -3826,11 +3849,14 @@ int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *, module_assert_mutex(); list_for_each_entry(mod, &modules, list) { + /* We hold module_mutex: no need for rcu_dereference_sched */ + struct mod_kallsyms *kallsyms = mod->kallsyms; + if (mod->state == MODULE_STATE_UNFORMED) continue; - for (i = 0; i < mod->num_symtab; i++) { - ret = fn(data, symname(mod, i), - mod, mod->symtab[i].st_value); + for (i = 0; i < kallsyms->num_symtab; i++) { + ret = fn(data, symname(kallsyms, i), + mod, kallsyms->symtab[i].st_value); if (ret != 0) return ret; } -- cgit From 0fb5b1fb30fba3671dd5b1489d78e93e08d62e4e Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Thu, 4 Feb 2016 00:52:12 -0500 Subject: block/sd: Return -EREMOTEIO when WRITE SAME and DISCARD are disabled When a storage device rejects a WRITE SAME command we will disable write same functionality for the device and return -EREMOTEIO to the block layer. -EREMOTEIO will in turn prevent DM from retrying the I/O and/or failing the path. Yiwen Jiang discovered a small race where WRITE SAME requests issued simultaneously would cause -EIO to be returned. This happened because any requests being prepared after WRITE SAME had been disabled for the device caused us to return BLKPREP_KILL. The latter caused the block layer to return -EIO upon completion. To overcome this we introduce BLKPREP_INVALID which indicates that this is an invalid request for the device. blk_peek_request() is modified to return -EREMOTEIO in that case. Reported-by: Yiwen Jiang Suggested-by: Mike Snitzer Reviewed-by: Hannes Reinicke Reviewed-by: Ewan Milne Reviewed-by: Yiwen Jiang Signed-off-by: Martin K. Petersen --- block/blk-core.c | 6 ++++-- drivers/scsi/sd.c | 4 ++-- include/linux/blkdev.h | 9 ++++++--- 3 files changed, 12 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/block/blk-core.c b/block/blk-core.c index 476244d59309..35607dd0223b 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2447,14 +2447,16 @@ struct request *blk_peek_request(struct request_queue *q) rq = NULL; break; - } else if (ret == BLKPREP_KILL) { + } else if (ret == BLKPREP_KILL || ret == BLKPREP_INVALID) { + int err = (ret == BLKPREP_INVALID) ? -EREMOTEIO : -EIO; + rq->cmd_flags |= REQ_QUIET; /* * Mark this request as started so we don't trigger * any debug logic in the end I/O path. */ blk_start_request(rq); - __blk_end_request_all(rq, -EIO); + __blk_end_request_all(rq, err); } else { printk(KERN_ERR "%s: bad return=%d\n", __func__, ret); break; diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index ec163d08f6c3..6e841c6da632 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -761,7 +761,7 @@ static int sd_setup_discard_cmnd(struct scsi_cmnd *cmd) break; default: - ret = BLKPREP_KILL; + ret = BLKPREP_INVALID; goto out; } @@ -839,7 +839,7 @@ static int sd_setup_write_same_cmnd(struct scsi_cmnd *cmd) int ret; if (sdkp->device->no_write_same) - return BLKPREP_KILL; + return BLKPREP_INVALID; BUG_ON(bio_offset(bio) || bio_iovec(bio).bv_len != sdp->sector_size); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index d372ea87ead5..a9b643a0647f 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -681,9 +681,12 @@ static inline bool blk_write_same_mergeable(struct bio *a, struct bio *b) /* * q->prep_rq_fn return values */ -#define BLKPREP_OK 0 /* serve it */ -#define BLKPREP_KILL 1 /* fatal error, kill */ -#define BLKPREP_DEFER 2 /* leave on queue */ +enum { + BLKPREP_OK, /* serve it */ + BLKPREP_KILL, /* fatal error, kill, return -EIO */ + BLKPREP_DEFER, /* leave on queue */ + BLKPREP_INVALID, /* invalid command, kill, return -EREMOTEIO */ +}; extern unsigned long blk_max_low_pfn, blk_max_pfn; -- cgit From 1f55c718c290616889c04946864a13ef30f64929 Mon Sep 17 00:00:00 2001 From: "Herton R. Krzesinski" Date: Thu, 14 Jan 2016 17:56:58 -0200 Subject: pty: make sure super_block is still valid in final /dev/tty close Considering current pty code and multiple devpts instances, it's possible to umount a devpts file system while a program still has /dev/tty opened pointing to a previosuly closed pty pair in that instance. In the case all ptmx and pts/N files are closed, umount can be done. If the program closes /dev/tty after umount is done, devpts_kill_index will use now an invalid super_block, which was already destroyed in the umount operation after running ->kill_sb. This is another "use after free" type of issue, but now related to the allocated super_block instance. To avoid the problem (warning at ida_remove and potential crashes) for this specific case, I added two functions in devpts which grabs additional references to the super_block, which pty code now uses so it makes sure the super block structure is still valid until pty shutdown is done. I also moved the additional inode references to the same functions, which also covered similar case with inode being freed before /dev/tty final close/shutdown. Signed-off-by: Herton R. Krzesinski Cc: stable@vger.kernel.org # 2.6.29+ Reviewed-by: Peter Hurley Signed-off-by: Greg Kroah-Hartman --- drivers/tty/pty.c | 9 ++++++--- fs/devpts/inode.c | 20 ++++++++++++++++++++ include/linux/devpts_fs.h | 4 ++++ 3 files changed, 30 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/drivers/tty/pty.c b/drivers/tty/pty.c index 3b5cde833ee5..2348fa613707 100644 --- a/drivers/tty/pty.c +++ b/drivers/tty/pty.c @@ -688,7 +688,7 @@ static void pty_unix98_shutdown(struct tty_struct *tty) else ptmx_inode = tty->link->driver_data; devpts_kill_index(ptmx_inode, tty->index); - iput(ptmx_inode); /* drop reference we acquired at ptmx_open */ + devpts_del_ref(ptmx_inode); } static const struct tty_operations ptm_unix98_ops = { @@ -785,9 +785,12 @@ static int ptmx_open(struct inode *inode, struct file *filp) * still have /dev/tty opened pointing to the master/slave pair (ptmx * is closed/released before /dev/tty), we must make sure that the inode * is still valid when we call the final pty_unix98_shutdown, thus we - * hold an additional reference to the ptmx inode + * hold an additional reference to the ptmx inode. For the same /dev/tty + * last close case, we also need to make sure the super_block isn't + * destroyed (devpts instance unmounted), before /dev/tty is closed and + * on its release devpts_kill_index is called. */ - ihold(inode); + devpts_add_ref(inode); tty_add_file(tty, filp); diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 1f107fd51328..655f21f99160 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -575,6 +575,26 @@ void devpts_kill_index(struct inode *ptmx_inode, int idx) mutex_unlock(&allocated_ptys_lock); } +/* + * pty code needs to hold extra references in case of last /dev/tty close + */ + +void devpts_add_ref(struct inode *ptmx_inode) +{ + struct super_block *sb = pts_sb_from_inode(ptmx_inode); + + atomic_inc(&sb->s_active); + ihold(ptmx_inode); +} + +void devpts_del_ref(struct inode *ptmx_inode) +{ + struct super_block *sb = pts_sb_from_inode(ptmx_inode); + + iput(ptmx_inode); + deactivate_super(sb); +} + /** * devpts_pty_new -- create a new inode in /dev/pts/ * @ptmx_inode: inode of the master diff --git a/include/linux/devpts_fs.h b/include/linux/devpts_fs.h index 251a2090a554..e0ee0b3000b2 100644 --- a/include/linux/devpts_fs.h +++ b/include/linux/devpts_fs.h @@ -19,6 +19,8 @@ int devpts_new_index(struct inode *ptmx_inode); void devpts_kill_index(struct inode *ptmx_inode, int idx); +void devpts_add_ref(struct inode *ptmx_inode); +void devpts_del_ref(struct inode *ptmx_inode); /* mknod in devpts */ struct inode *devpts_pty_new(struct inode *ptmx_inode, dev_t device, int index, void *priv); @@ -32,6 +34,8 @@ void devpts_pty_kill(struct inode *inode); /* Dummy stubs in the no-pty case */ static inline int devpts_new_index(struct inode *ptmx_inode) { return -EINVAL; } static inline void devpts_kill_index(struct inode *ptmx_inode, int idx) { } +static inline void devpts_add_ref(struct inode *ptmx_inode) { } +static inline void devpts_del_ref(struct inode *ptmx_inode) { } static inline struct inode *devpts_pty_new(struct inode *ptmx_inode, dev_t device, int index, void *priv) { -- cgit From 7047f17d70fc0599563d30d0791692cb5fe42ae6 Mon Sep 17 00:00:00 2001 From: "K. Y. Srinivasan" Date: Fri, 25 Dec 2015 20:00:30 -0800 Subject: Drivers: hv: vmbus: Add vendor and device atttributes Add vendor and device attributes to VMBUS devices. These will be used by Hyper-V tools as well user-level RDMA libraries that will use the vendor/device tuple to discover the RDMA device. Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- Documentation/ABI/stable/sysfs-bus-vmbus | 14 +++ drivers/hv/channel_mgmt.c | 166 +++++++++++++++++++++++-------- drivers/hv/vmbus_drv.c | 21 ++++ include/linux/hyperv.h | 28 ++++++ 4 files changed, 186 insertions(+), 43 deletions(-) (limited to 'include/linux') diff --git a/Documentation/ABI/stable/sysfs-bus-vmbus b/Documentation/ABI/stable/sysfs-bus-vmbus index 636e938d5e33..5d0125f7bcaf 100644 --- a/Documentation/ABI/stable/sysfs-bus-vmbus +++ b/Documentation/ABI/stable/sysfs-bus-vmbus @@ -27,3 +27,17 @@ Description: The mapping of which primary/sub channels are bound to which Virtual Processors. Format: Users: tools/hv/lsvmbus + +What: /sys/bus/vmbus/devices/vmbus_*/device +Date: Dec. 2015 +KernelVersion: 4.5 +Contact: K. Y. Srinivasan +Description: The 16 bit device ID of the device +Users: tools/hv/lsvmbus and user level RDMA libraries + +What: /sys/bus/vmbus/devices/vmbus_*/vendor +Date: Dec. 2015 +KernelVersion: 4.5 +Contact: K. Y. Srinivasan +Description: The 16 bit vendor ID of the device +Users: tools/hv/lsvmbus and user level RDMA libraries diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c index 1c1ad47042c5..107d72f9834d 100644 --- a/drivers/hv/channel_mgmt.c +++ b/drivers/hv/channel_mgmt.c @@ -32,8 +32,122 @@ #include "hyperv_vmbus.h" -static void init_vp_index(struct vmbus_channel *channel, - const uuid_le *type_guid); +static void init_vp_index(struct vmbus_channel *channel, u16 dev_type); + +static const struct vmbus_device vmbus_devs[] = { + /* IDE */ + { .dev_type = HV_IDE, + HV_IDE_GUID, + .perf_device = true, + }, + + /* SCSI */ + { .dev_type = HV_SCSI, + HV_SCSI_GUID, + .perf_device = true, + }, + + /* Fibre Channel */ + { .dev_type = HV_FC, + HV_SYNTHFC_GUID, + .perf_device = true, + }, + + /* Synthetic NIC */ + { .dev_type = HV_NIC, + HV_NIC_GUID, + .perf_device = true, + }, + + /* Network Direct */ + { .dev_type = HV_ND, + HV_ND_GUID, + .perf_device = true, + }, + + /* PCIE */ + { .dev_type = HV_PCIE, + HV_PCIE_GUID, + .perf_device = true, + }, + + /* Synthetic Frame Buffer */ + { .dev_type = HV_FB, + HV_SYNTHVID_GUID, + .perf_device = false, + }, + + /* Synthetic Keyboard */ + { .dev_type = HV_KBD, + HV_KBD_GUID, + .perf_device = false, + }, + + /* Synthetic MOUSE */ + { .dev_type = HV_MOUSE, + HV_MOUSE_GUID, + .perf_device = false, + }, + + /* KVP */ + { .dev_type = HV_KVP, + HV_KVP_GUID, + .perf_device = false, + }, + + /* Time Synch */ + { .dev_type = HV_TS, + HV_TS_GUID, + .perf_device = false, + }, + + /* Heartbeat */ + { .dev_type = HV_HB, + HV_HEART_BEAT_GUID, + .perf_device = false, + }, + + /* Shutdown */ + { .dev_type = HV_SHUTDOWN, + HV_SHUTDOWN_GUID, + .perf_device = false, + }, + + /* File copy */ + { .dev_type = HV_FCOPY, + HV_FCOPY_GUID, + .perf_device = false, + }, + + /* Backup */ + { .dev_type = HV_BACKUP, + HV_VSS_GUID, + .perf_device = false, + }, + + /* Dynamic Memory */ + { .dev_type = HV_DM, + HV_DM_GUID, + .perf_device = false, + }, + + /* Unknown GUID */ + { .dev_type = HV_UNKOWN, + .perf_device = false, + }, +}; + +static u16 hv_get_dev_type(const uuid_le *guid) +{ + u16 i; + + for (i = HV_IDE; i < HV_UNKOWN; i++) { + if (!uuid_le_cmp(*guid, vmbus_devs[i].guid)) + return i; + } + pr_info("Unknown GUID: %pUl\n", guid); + return i; +} /** * vmbus_prep_negotiate_resp() - Create default response for Hyper-V Negotiate message @@ -251,6 +365,7 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel) struct vmbus_channel *channel; bool fnew = true; unsigned long flags; + u16 dev_type; /* Make sure this is a new offer */ mutex_lock(&vmbus_connection.channel_mutex); @@ -288,7 +403,9 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel) goto err_free_chan; } - init_vp_index(newchannel, &newchannel->offermsg.offer.if_type); + dev_type = hv_get_dev_type(&newchannel->offermsg.offer.if_type); + + init_vp_index(newchannel, dev_type); if (newchannel->target_cpu != get_cpu()) { put_cpu(); @@ -325,6 +442,7 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel) if (!newchannel->device_obj) goto err_deq_chan; + newchannel->device_obj->device_id = dev_type; /* * Add the new device to the bus. This will kick off device-driver * binding which eventually invokes the device driver's AddDevice() @@ -358,37 +476,6 @@ err_free_chan: free_channel(newchannel); } -enum { - IDE = 0, - SCSI, - FC, - NIC, - ND_NIC, - PCIE, - MAX_PERF_CHN, -}; - -/* - * This is an array of device_ids (device types) that are performance critical. - * We attempt to distribute the interrupt load for these devices across - * all available CPUs. - */ -static const struct hv_vmbus_device_id hp_devs[] = { - /* IDE */ - { HV_IDE_GUID, }, - /* Storage - SCSI */ - { HV_SCSI_GUID, }, - /* Storage - FC */ - { HV_SYNTHFC_GUID, }, - /* Network */ - { HV_NIC_GUID, }, - /* NetworkDirect Guest RDMA */ - { HV_ND_GUID, }, - /* PCI Express Pass Through */ - { HV_PCIE_GUID, }, -}; - - /* * We use this state to statically distribute the channel interrupt load. */ @@ -405,22 +492,15 @@ static int next_numa_node_id; * For pre-win8 hosts or non-performance critical channels we assign the * first CPU in the first NUMA node. */ -static void init_vp_index(struct vmbus_channel *channel, const uuid_le *type_guid) +static void init_vp_index(struct vmbus_channel *channel, u16 dev_type) { u32 cur_cpu; - int i; - bool perf_chn = false; + bool perf_chn = vmbus_devs[dev_type].perf_device; struct vmbus_channel *primary = channel->primary_channel; int next_node; struct cpumask available_mask; struct cpumask *alloced_mask; - for (i = IDE; i < MAX_PERF_CHN; i++) { - if (!uuid_le_cmp(*type_guid, hp_devs[i].guid)) { - perf_chn = true; - break; - } - } if ((vmbus_proto_version == VERSION_WS2008) || (vmbus_proto_version == VERSION_WIN7) || (!perf_chn)) { /* diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c index 328e4c3808e0..3668a95778ec 100644 --- a/drivers/hv/vmbus_drv.c +++ b/drivers/hv/vmbus_drv.c @@ -477,6 +477,24 @@ static ssize_t channel_vp_mapping_show(struct device *dev, } static DEVICE_ATTR_RO(channel_vp_mapping); +static ssize_t vendor_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + return sprintf(buf, "0x%x\n", hv_dev->vendor_id); +} +static DEVICE_ATTR_RO(vendor); + +static ssize_t device_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + return sprintf(buf, "0x%x\n", hv_dev->device_id); +} +static DEVICE_ATTR_RO(device); + /* Set up per device attributes in /sys/bus/vmbus/devices/ */ static struct attribute *vmbus_attrs[] = { &dev_attr_id.attr, @@ -502,6 +520,8 @@ static struct attribute *vmbus_attrs[] = { &dev_attr_in_read_bytes_avail.attr, &dev_attr_in_write_bytes_avail.attr, &dev_attr_channel_vp_mapping.attr, + &dev_attr_vendor.attr, + &dev_attr_device.attr, NULL, }; ATTRIBUTE_GROUPS(vmbus); @@ -957,6 +977,7 @@ struct hv_device *vmbus_device_create(const uuid_le *type, memcpy(&child_device_obj->dev_type, type, sizeof(uuid_le)); memcpy(&child_device_obj->dev_instance, instance, sizeof(uuid_le)); + child_device_obj->vendor_id = 0x1414; /* MSFT vendor ID */ return child_device_obj; diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 753dbad0bf94..3172521df805 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -633,6 +633,32 @@ enum hv_signal_policy { HV_SIGNAL_POLICY_EXPLICIT, }; +enum vmbus_device_type { + HV_IDE = 0, + HV_SCSI, + HV_FC, + HV_NIC, + HV_ND, + HV_PCIE, + HV_FB, + HV_KBD, + HV_MOUSE, + HV_KVP, + HV_TS, + HV_HB, + HV_SHUTDOWN, + HV_FCOPY, + HV_BACKUP, + HV_DM, + HV_UNKOWN, +}; + +struct vmbus_device { + u16 dev_type; + uuid_le guid; + bool perf_device; +}; + struct vmbus_channel { /* Unique channel id */ int id; @@ -959,6 +985,8 @@ struct hv_device { /* the device instance id of this device */ uuid_le dev_instance; + u16 vendor_id; + u16 device_id; struct device device; -- cgit From 3c75354d043ad546148d6992e40033ecaefc5ea5 Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:37 -0800 Subject: Drivers: hv: vmbus: add a helper function to set a channel's pending send size This will be used by the coming net/hvsock driver. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- include/linux/hyperv.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include/linux') diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 3172521df805..4af51a3f98d9 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -816,6 +816,12 @@ static inline void *get_per_channel_state(struct vmbus_channel *c) return c->per_channel_state; } +static inline void set_channel_pending_send_size(struct vmbus_channel *c, + u32 size) +{ + c->outbound.ring_buffer->pending_send_sz = size; +} + void vmbus_onmessage(void *context); int vmbus_request_offers(void); -- cgit From e8d6ca023efce3bd80050dcd9e708ee3cf8babd4 Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:38 -0800 Subject: Drivers: hv: vmbus: define the new offer type for Hyper-V socket (hvsock) A helper function is also added. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- include/linux/hyperv.h | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'include/linux') diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 4af51a3f98d9..79c4aa70eff3 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -235,6 +235,7 @@ struct vmbus_channel_offer { #define VMBUS_CHANNEL_LOOPBACK_OFFER 0x100 #define VMBUS_CHANNEL_PARENT_OFFER 0x200 #define VMBUS_CHANNEL_REQUEST_MONITORED_NOTIFICATION 0x400 +#define VMBUS_CHANNEL_TLNPI_PROVIDER_OFFER 0x2000 struct vmpacket_descriptor { u16 type; @@ -795,6 +796,12 @@ struct vmbus_channel { enum hv_signal_policy signal_policy; }; +static inline bool is_hvsock_channel(const struct vmbus_channel *c) +{ + return !!(c->offermsg.offer.chn_flags & + VMBUS_CHANNEL_TLNPI_PROVIDER_OFFER); +} + static inline void set_channel_signal_state(struct vmbus_channel *c, enum hv_signal_policy policy) { -- cgit From 5c23a1a5c60b0f472cfa61cd7d8279f8aaeb5b64 Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:40 -0800 Subject: Drivers: hv: vmbus: define a new VMBus message type for hvsock A function to send the type of message is also added. The coming net/hvsock driver will use this function to proactively request the host to offer a VMBus channel for a new hvsock connection. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- drivers/hv/channel.c | 15 +++++++++++++++ drivers/hv/channel_mgmt.c | 4 ++++ include/linux/hyperv.h | 13 +++++++++++++ 3 files changed, 32 insertions(+) (limited to 'include/linux') diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c index 3f0453302146..fcab234796ef 100644 --- a/drivers/hv/channel.c +++ b/drivers/hv/channel.c @@ -219,6 +219,21 @@ error0: } EXPORT_SYMBOL_GPL(vmbus_open); +/* Used for Hyper-V Socket: a guest client's connect() to the host */ +int vmbus_send_tl_connect_request(const uuid_le *shv_guest_servie_id, + const uuid_le *shv_host_servie_id) +{ + struct vmbus_channel_tl_connect_request conn_msg; + + memset(&conn_msg, 0, sizeof(conn_msg)); + conn_msg.header.msgtype = CHANNELMSG_TL_CONNECT_REQUEST; + conn_msg.guest_endpoint_id = *shv_guest_servie_id; + conn_msg.host_service_id = *shv_host_servie_id; + + return vmbus_post_msg(&conn_msg, sizeof(conn_msg)); +} +EXPORT_SYMBOL_GPL(vmbus_send_tl_connect_request); + /* * create_gpadl_header - Creates a gpadl for the specified buffer */ diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c index d6c611457601..60ca25b93b4c 100644 --- a/drivers/hv/channel_mgmt.c +++ b/drivers/hv/channel_mgmt.c @@ -958,6 +958,10 @@ struct vmbus_channel_message_table_entry {CHANNELMSG_VERSION_RESPONSE, 1, vmbus_onversion_response}, {CHANNELMSG_UNLOAD, 0, NULL}, {CHANNELMSG_UNLOAD_RESPONSE, 1, vmbus_unload_response}, + {CHANNELMSG_18, 0, NULL}, + {CHANNELMSG_19, 0, NULL}, + {CHANNELMSG_20, 0, NULL}, + {CHANNELMSG_TL_CONNECT_REQUEST, 0, NULL}, }; /* diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 79c4aa70eff3..898eac9e8c13 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -392,6 +392,10 @@ enum vmbus_channel_message_type { CHANNELMSG_VERSION_RESPONSE = 15, CHANNELMSG_UNLOAD = 16, CHANNELMSG_UNLOAD_RESPONSE = 17, + CHANNELMSG_18 = 18, + CHANNELMSG_19 = 19, + CHANNELMSG_20 = 20, + CHANNELMSG_TL_CONNECT_REQUEST = 21, CHANNELMSG_COUNT }; @@ -562,6 +566,13 @@ struct vmbus_channel_initiate_contact { u64 monitor_page2; } __packed; +/* Hyper-V socket: guest's connect()-ing to host */ +struct vmbus_channel_tl_connect_request { + struct vmbus_channel_message_header header; + uuid_le guest_endpoint_id; + uuid_le host_service_id; +} __packed; + struct vmbus_channel_version_response { struct vmbus_channel_message_header header; u8 version_supported; @@ -1283,4 +1294,6 @@ void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid); extern __u32 vmbus_proto_version; +int vmbus_send_tl_connect_request(const uuid_le *shv_guest_servie_id, + const uuid_le *shv_host_servie_id); #endif /* _HYPERV_H */ -- cgit From 8981da320a11217589aa3c50f9e891bcdef07ece Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:41 -0800 Subject: Drivers: hv: vmbus: add a hvsock flag in struct hv_driver Only the coming hv_sock driver has a "true" value for this flag. We treat the hvsock offers/channels as special VMBus devices. Since the hv_sock driver handles all the hvsock offers/channels, we need to tweak vmbus_match() for hv_sock driver, so we introduce this flag. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- drivers/hv/vmbus_drv.c | 4 ++++ include/linux/hyperv.h | 14 ++++++++++++++ 2 files changed, 18 insertions(+) (limited to 'include/linux') diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c index 3668a95778ec..063e5f53ca78 100644 --- a/drivers/hv/vmbus_drv.c +++ b/drivers/hv/vmbus_drv.c @@ -582,6 +582,10 @@ static int vmbus_match(struct device *device, struct device_driver *driver) struct hv_driver *drv = drv_to_hv_drv(driver); struct hv_device *hv_dev = device_to_hv_device(device); + /* The hv_sock driver handles all hv_sock offers. */ + if (is_hvsock_channel(hv_dev->channel)) + return drv->hvsock; + if (hv_vmbus_get_id(drv->id_table, &hv_dev->dev_type)) return 1; diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 898eac9e8c13..f636f91f104b 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -990,6 +990,20 @@ extern void vmbus_ontimer(unsigned long data); struct hv_driver { const char *name; + /* + * A hvsock offer, which has a VMBUS_CHANNEL_TLNPI_PROVIDER_OFFER + * channel flag, actually doesn't mean a synthetic device because the + * offer's if_type/if_instance can change for every new hvsock + * connection. + * + * However, to facilitate the notification of new-offer/rescind-offer + * from vmbus driver to hvsock driver, we can handle hvsock offer as + * a special vmbus device, and hence we need the below flag to + * indicate if the driver is the hvsock driver or not: we need to + * specially treat the hvosck offer & driver in vmbus_match(). + */ + bool hvsock; + /* the device type supported by this driver */ uuid_le dev_type; const struct hv_vmbus_device_id *id_table; -- cgit From 499e8401a515d04daa986b995da710d2b9737764 Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:42 -0800 Subject: Drivers: hv: vmbus: add a per-channel rescind callback This will be used by the coming hv_sock driver. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- drivers/hv/channel_mgmt.c | 11 +++++++++++ include/linux/hyperv.h | 9 +++++++++ 2 files changed, 20 insertions(+) (limited to 'include/linux') diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c index 60ca25b93b4c..76864c98a110 100644 --- a/drivers/hv/channel_mgmt.c +++ b/drivers/hv/channel_mgmt.c @@ -741,6 +741,10 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) spin_unlock_irqrestore(&channel->lock, flags); if (channel->device_obj) { + if (channel->chn_rescind_callback) { + channel->chn_rescind_callback(channel); + return; + } /* * We will have to unregister this device from the * driver core. @@ -1110,3 +1114,10 @@ bool vmbus_are_subchannels_present(struct vmbus_channel *primary) return ret; } EXPORT_SYMBOL_GPL(vmbus_are_subchannels_present); + +void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, + void (*chn_rescind_cb)(struct vmbus_channel *)) +{ + channel->chn_rescind_callback = chn_rescind_cb; +} +EXPORT_SYMBOL_GPL(vmbus_set_chn_rescind_callback); diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index f636f91f104b..2e54e34e5feb 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -765,6 +765,12 @@ struct vmbus_channel { */ void (*sc_creation_callback)(struct vmbus_channel *new_sc); + /* + * Channel rescind callback. Some channels (the hvsock ones), need to + * register a callback which is invoked in vmbus_onoffer_rescind(). + */ + void (*chn_rescind_callback)(struct vmbus_channel *channel); + /* * The spinlock to protect the structure. It is being used to protect * test-and-set access to various attributes of the structure as well @@ -851,6 +857,9 @@ int vmbus_request_offers(void); void vmbus_set_sc_create_callback(struct vmbus_channel *primary_channel, void (*sc_cr_cb)(struct vmbus_channel *new_sc)); +void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, + void (*chn_rescind_cb)(struct vmbus_channel *)); + /* * Retrieve the (sub) channel on which to send an outgoing request. * When a primary channel has multiple sub-channels, we choose a -- cgit From 85d9aa705184a4504d0330017e3956fcdae8a9d6 Mon Sep 17 00:00:00 2001 From: Dexuan Cui Date: Wed, 27 Jan 2016 22:29:43 -0800 Subject: Drivers: hv: vmbus: add an API vmbus_hvsock_device_unregister() The hvsock driver needs this API to release all the resources related to the channel. Signed-off-by: Dexuan Cui Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- drivers/hv/channel_mgmt.c | 33 ++++++++++++++++++++++++++++----- drivers/hv/connection.c | 4 ++-- include/linux/hyperv.h | 2 ++ 3 files changed, 32 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c index 76864c98a110..cf311be88cb4 100644 --- a/drivers/hv/channel_mgmt.c +++ b/drivers/hv/channel_mgmt.c @@ -310,6 +310,7 @@ void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid) vmbus_release_relid(relid); BUG_ON(!channel->rescind); + BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex)); if (channel->target_cpu != get_cpu()) { put_cpu(); @@ -321,9 +322,7 @@ void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid) } if (channel->primary_channel == NULL) { - mutex_lock(&vmbus_connection.channel_mutex); list_del(&channel->listentry); - mutex_unlock(&vmbus_connection.channel_mutex); primary_channel = channel; } else { @@ -367,6 +366,7 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel) bool fnew = true; unsigned long flags; u16 dev_type; + int ret; /* Make sure this is a new offer */ mutex_lock(&vmbus_connection.channel_mutex); @@ -449,7 +449,11 @@ static void vmbus_process_offer(struct vmbus_channel *newchannel) * binding which eventually invokes the device driver's AddDevice() * method. */ - if (vmbus_device_register(newchannel->device_obj) != 0) { + mutex_lock(&vmbus_connection.channel_mutex); + ret = vmbus_device_register(newchannel->device_obj); + mutex_unlock(&vmbus_connection.channel_mutex); + + if (ret != 0) { pr_err("unable to add child device object (relid %d)\n", newchannel->offermsg.child_relid); kfree(newchannel->device_obj); @@ -725,6 +729,8 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) struct device *dev; rescind = (struct vmbus_channel_rescind_offer *)hdr; + + mutex_lock(&vmbus_connection.channel_mutex); channel = relid2channel(rescind->child_relid); if (channel == NULL) { @@ -733,7 +739,7 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) * vmbus_process_offer(), we have already invoked * vmbus_release_relid() on error. */ - return; + goto out; } spin_lock_irqsave(&channel->lock, flags); @@ -743,7 +749,7 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) if (channel->device_obj) { if (channel->chn_rescind_callback) { channel->chn_rescind_callback(channel); - return; + goto out; } /* * We will have to unregister this device from the @@ -758,8 +764,25 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) hv_process_channel_removal(channel, channel->offermsg.child_relid); } + +out: + mutex_unlock(&vmbus_connection.channel_mutex); } +void vmbus_hvsock_device_unregister(struct vmbus_channel *channel) +{ + mutex_lock(&vmbus_connection.channel_mutex); + + BUG_ON(!is_hvsock_channel(channel)); + + channel->rescind = true; + vmbus_device_unregister(channel->device_obj); + + mutex_unlock(&vmbus_connection.channel_mutex); +} +EXPORT_SYMBOL_GPL(vmbus_hvsock_device_unregister); + + /* * vmbus_onoffers_delivered - * This is invoked when all offers have been delivered. diff --git a/drivers/hv/connection.c b/drivers/hv/connection.c index 4a320e60641a..fa86b2cb28b8 100644 --- a/drivers/hv/connection.c +++ b/drivers/hv/connection.c @@ -288,7 +288,8 @@ struct vmbus_channel *relid2channel(u32 relid) struct list_head *cur, *tmp; struct vmbus_channel *cur_sc; - mutex_lock(&vmbus_connection.channel_mutex); + BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex)); + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { if (channel->offermsg.child_relid == relid) { found_channel = channel; @@ -307,7 +308,6 @@ struct vmbus_channel *relid2channel(u32 relid) } } } - mutex_unlock(&vmbus_connection.channel_mutex); return found_channel; } diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index 2e54e34e5feb..c056f058dcf8 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -1069,6 +1069,8 @@ int __must_check __vmbus_driver_register(struct hv_driver *hv_driver, const char *mod_name); void vmbus_driver_unregister(struct hv_driver *hv_driver); +void vmbus_hvsock_device_unregister(struct vmbus_channel *channel); + int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj, resource_size_t min, resource_size_t max, resource_size_t size, resource_size_t align, -- cgit From fe760e4d64fe5c17c39e86c410d41f6587ee88bc Mon Sep 17 00:00:00 2001 From: "K. Y. Srinivasan" Date: Wed, 27 Jan 2016 22:29:45 -0800 Subject: Drivers: hv: vmbus: Give control over how the ring access is serialized On the channel send side, many of the VMBUS device drivers explicity serialize access to the outgoing ring buffer. Give more control to the VMBUS device drivers in terms how to serialize accesss to the outgoing ring buffer. The default behavior will be to aquire the ring lock to preserve the current behavior. Signed-off-by: K. Y. Srinivasan Signed-off-by: Greg Kroah-Hartman --- drivers/hv/channel.c | 15 +++++++++++---- drivers/hv/channel_mgmt.c | 1 + drivers/hv/hyperv_vmbus.h | 2 +- drivers/hv/ring_buffer.c | 13 ++++++++----- include/linux/hyperv.h | 16 ++++++++++++++++ 5 files changed, 37 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c index fcab234796ef..56dd261f7142 100644 --- a/drivers/hv/channel.c +++ b/drivers/hv/channel.c @@ -639,6 +639,7 @@ int vmbus_sendpacket_ctl(struct vmbus_channel *channel, void *buffer, u64 aligned_data = 0; int ret; bool signal = false; + bool lock = channel->acquire_ring_lock; int num_vecs = ((bufferlen != 0) ? 3 : 1); @@ -658,7 +659,7 @@ int vmbus_sendpacket_ctl(struct vmbus_channel *channel, void *buffer, bufferlist[2].iov_len = (packetlen_aligned - packetlen); ret = hv_ringbuffer_write(&channel->outbound, bufferlist, num_vecs, - &signal); + &signal, lock); /* * Signalling the host is conditional on many factors: @@ -738,6 +739,7 @@ int vmbus_sendpacket_pagebuffer_ctl(struct vmbus_channel *channel, struct kvec bufferlist[3]; u64 aligned_data = 0; bool signal = false; + bool lock = channel->acquire_ring_lock; if (pagecount > MAX_PAGE_BUFFER_COUNT) return -EINVAL; @@ -774,7 +776,8 @@ int vmbus_sendpacket_pagebuffer_ctl(struct vmbus_channel *channel, bufferlist[2].iov_base = &aligned_data; bufferlist[2].iov_len = (packetlen_aligned - packetlen); - ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, + &signal, lock); /* * Signalling the host is conditional on many factors: @@ -837,6 +840,7 @@ int vmbus_sendpacket_mpb_desc(struct vmbus_channel *channel, struct kvec bufferlist[3]; u64 aligned_data = 0; bool signal = false; + bool lock = channel->acquire_ring_lock; packetlen = desc_size + bufferlen; packetlen_aligned = ALIGN(packetlen, sizeof(u64)); @@ -856,7 +860,8 @@ int vmbus_sendpacket_mpb_desc(struct vmbus_channel *channel, bufferlist[2].iov_base = &aligned_data; bufferlist[2].iov_len = (packetlen_aligned - packetlen); - ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, + &signal, lock); if (ret == 0 && signal) vmbus_setevent(channel); @@ -881,6 +886,7 @@ int vmbus_sendpacket_multipagebuffer(struct vmbus_channel *channel, struct kvec bufferlist[3]; u64 aligned_data = 0; bool signal = false; + bool lock = channel->acquire_ring_lock; u32 pfncount = NUM_PAGES_SPANNED(multi_pagebuffer->offset, multi_pagebuffer->len); @@ -919,7 +925,8 @@ int vmbus_sendpacket_multipagebuffer(struct vmbus_channel *channel, bufferlist[2].iov_base = &aligned_data; bufferlist[2].iov_len = (packetlen_aligned - packetlen); - ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, + &signal, lock); if (ret == 0 && signal) vmbus_setevent(channel); diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c index cf311be88cb4..b40f429aaa13 100644 --- a/drivers/hv/channel_mgmt.c +++ b/drivers/hv/channel_mgmt.c @@ -259,6 +259,7 @@ static struct vmbus_channel *alloc_channel(void) return NULL; channel->id = atomic_inc_return(&chan_num); + channel->acquire_ring_lock = true; spin_lock_init(&channel->inbound_lock); spin_lock_init(&channel->lock); diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h index ac7aa303c37d..b9ea7f59036b 100644 --- a/drivers/hv/hyperv_vmbus.h +++ b/drivers/hv/hyperv_vmbus.h @@ -529,7 +529,7 @@ void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info); int hv_ringbuffer_write(struct hv_ring_buffer_info *ring_info, struct kvec *kv_list, - u32 kv_count, bool *signal); + u32 kv_count, bool *signal, bool lock); int hv_ringbuffer_read(struct hv_ring_buffer_info *inring_info, void *buffer, u32 buflen, u32 *buffer_actual_len, diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c index 1145f3b8e4e0..5613e2b5cff7 100644 --- a/drivers/hv/ring_buffer.c +++ b/drivers/hv/ring_buffer.c @@ -314,7 +314,7 @@ void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info) /* Write to the ring buffer. */ int hv_ringbuffer_write(struct hv_ring_buffer_info *outring_info, - struct kvec *kv_list, u32 kv_count, bool *signal) + struct kvec *kv_list, u32 kv_count, bool *signal, bool lock) { int i = 0; u32 bytes_avail_towrite; @@ -324,14 +324,15 @@ int hv_ringbuffer_write(struct hv_ring_buffer_info *outring_info, u32 next_write_location; u32 old_write; u64 prev_indices = 0; - unsigned long flags; + unsigned long flags = 0; for (i = 0; i < kv_count; i++) totalbytes_towrite += kv_list[i].iov_len; totalbytes_towrite += sizeof(u64); - spin_lock_irqsave(&outring_info->ring_lock, flags); + if (lock) + spin_lock_irqsave(&outring_info->ring_lock, flags); hv_get_ringbuffer_availbytes(outring_info, &bytes_avail_toread, @@ -343,7 +344,8 @@ int hv_ringbuffer_write(struct hv_ring_buffer_info *outring_info, * is empty since the read index == write index. */ if (bytes_avail_towrite <= totalbytes_towrite) { - spin_unlock_irqrestore(&outring_info->ring_lock, flags); + if (lock) + spin_unlock_irqrestore(&outring_info->ring_lock, flags); return -EAGAIN; } @@ -374,7 +376,8 @@ int hv_ringbuffer_write(struct hv_ring_buffer_info *outring_info, hv_set_next_write_location(outring_info, next_write_location); - spin_unlock_irqrestore(&outring_info->ring_lock, flags); + if (lock) + spin_unlock_irqrestore(&outring_info->ring_lock, flags); *signal = hv_need_to_signal(old_write, outring_info); return 0; diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h index c056f058dcf8..d23dab0d770b 100644 --- a/include/linux/hyperv.h +++ b/include/linux/hyperv.h @@ -811,8 +811,24 @@ struct vmbus_channel { * signaling control. */ enum hv_signal_policy signal_policy; + /* + * On the channel send side, many of the VMBUS + * device drivers explicity serialize access to the + * outgoing ring buffer. Give more control to the + * VMBUS device drivers in terms how to serialize + * accesss to the outgoing ring buffer. + * The default behavior will be to aquire the + * ring lock to preserve the current behavior. + */ + bool acquire_ring_lock; + }; +static inline void set_channel_lock_state(struct vmbus_channel *c, bool state) +{ + c->acquire_ring_lock = state; +} + static inline bool is_hvsock_channel(const struct vmbus_channel *c) { return !!(c->offermsg.offer.chn_flags & -- cgit From f42a0fd13bd281811e7457b28d939c8e8b808868 Mon Sep 17 00:00:00 2001 From: Jorgen Hansen Date: Thu, 12 Nov 2015 01:29:32 -0800 Subject: VMCI: Use 32bit atomics for queue headers on X86_32 This change restricts the reading and setting of the head and tail pointers on 32bit X86 to 32bit for both correctness and performance reasons. On uniprocessor X86_32, the atomic64_read may be implemented as a non-locked cmpxchg8b. This may result in updates to the pointers done by the VMCI device being overwritten. On MP systems, there is no such correctness issue, but using 32bit atomics avoids the overhead of the locked 64bit operation. All this is safe because the queue size on 32bit systems will never exceed a 32bit value. Reviewed-by: Thomas Hellstrom Signed-off-by: Jorgen Hansen Signed-off-by: Greg Kroah-Hartman --- drivers/misc/vmw_vmci/vmci_driver.c | 2 +- include/linux/vmw_vmci_defs.h | 43 +++++++++++++++++++++++++++++++++---- 2 files changed, 40 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/drivers/misc/vmw_vmci/vmci_driver.c b/drivers/misc/vmw_vmci/vmci_driver.c index b823f9a6e464..896be150e28f 100644 --- a/drivers/misc/vmw_vmci/vmci_driver.c +++ b/drivers/misc/vmw_vmci/vmci_driver.c @@ -113,5 +113,5 @@ module_exit(vmci_drv_exit); MODULE_AUTHOR("VMware, Inc."); MODULE_DESCRIPTION("VMware Virtual Machine Communication Interface."); -MODULE_VERSION("1.1.3.0-k"); +MODULE_VERSION("1.1.4.0-k"); MODULE_LICENSE("GPL v2"); diff --git a/include/linux/vmw_vmci_defs.h b/include/linux/vmw_vmci_defs.h index 65ac54c61c18..1bd31a38c51e 100644 --- a/include/linux/vmw_vmci_defs.h +++ b/include/linux/vmw_vmci_defs.h @@ -733,6 +733,41 @@ static inline void *vmci_event_data_payload(struct vmci_event_data *ev_data) return (void *)vmci_event_data_const_payload(ev_data); } +/* + * Helper to read a value from a head or tail pointer. For X86_32, the + * pointer is treated as a 32bit value, since the pointer value + * never exceeds a 32bit value in this case. Also, doing an + * atomic64_read on X86_32 uniprocessor systems may be implemented + * as a non locked cmpxchg8b, that may end up overwriting updates done + * by the VMCI device to the memory location. On 32bit SMP, the lock + * prefix will be used, so correctness isn't an issue, but using a + * 64bit operation still adds unnecessary overhead. + */ +static inline u64 vmci_q_read_pointer(atomic64_t *var) +{ +#if defined(CONFIG_X86_32) + return atomic_read((atomic_t *)var); +#else + return atomic64_read(var); +#endif +} + +/* + * Helper to set the value of a head or tail pointer. For X86_32, the + * pointer is treated as a 32bit value, since the pointer value + * never exceeds a 32bit value in this case. On 32bit SMP, using a + * locked cmpxchg8b adds unnecessary overhead. + */ +static inline void vmci_q_set_pointer(atomic64_t *var, + u64 new_val) +{ +#if defined(CONFIG_X86_32) + return atomic_set((atomic_t *)var, (u32)new_val); +#else + return atomic64_set(var, new_val); +#endif +} + /* * Helper to add a given offset to a head or tail pointer. Wraps the * value of the pointer around the max size of the queue. @@ -741,14 +776,14 @@ static inline void vmci_qp_add_pointer(atomic64_t *var, size_t add, u64 size) { - u64 new_val = atomic64_read(var); + u64 new_val = vmci_q_read_pointer(var); if (new_val >= size - add) new_val -= size; new_val += add; - atomic64_set(var, new_val); + vmci_q_set_pointer(var, new_val); } /* @@ -758,7 +793,7 @@ static inline u64 vmci_q_header_producer_tail(const struct vmci_queue_header *q_header) { struct vmci_queue_header *qh = (struct vmci_queue_header *)q_header; - return atomic64_read(&qh->producer_tail); + return vmci_q_read_pointer(&qh->producer_tail); } /* @@ -768,7 +803,7 @@ static inline u64 vmci_q_header_consumer_head(const struct vmci_queue_header *q_header) { struct vmci_queue_header *qh = (struct vmci_queue_header *)q_header; - return atomic64_read(&qh->consumer_head); + return vmci_q_read_pointer(&qh->consumer_head); } /* -- cgit From 52210c8745e418f82f3f0aeeee01d7bc4858812a Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Tue, 2 Feb 2016 14:14:01 -0700 Subject: coresight: implementing 'cpu_id()' API Other than plainly parsing the device tree there is no way to know which CPU a tracer is affined to. As such adding an interface to lookup the CPU field enclosed in the etm_drvdata structure that was initialised at boot time. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etm3x.c | 8 ++++++++ drivers/hwtracing/coresight/coresight-etm4x.c | 8 ++++++++ include/linux/coresight.h | 3 +++ 3 files changed, 19 insertions(+) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index 5981fcc69960..aae80e14508d 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -315,6 +315,13 @@ static void etm_enable_hw(void *info) dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu); } +static int etm_cpu_id(struct coresight_device *csdev) +{ + struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + return drvdata->cpu; +} + static int etm_trace_id(struct coresight_device *csdev) { struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); @@ -421,6 +428,7 @@ static void etm_disable(struct coresight_device *csdev) } static const struct coresight_ops_source etm_source_ops = { + .cpu_id = etm_cpu_id, .trace_id = etm_trace_id, .enable = etm_enable, .disable = etm_disable, diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c index 167004f9c42b..b6ae9cb6ff57 100644 --- a/drivers/hwtracing/coresight/coresight-etm4x.c +++ b/drivers/hwtracing/coresight/coresight-etm4x.c @@ -63,6 +63,13 @@ static bool etm4_arch_supported(u8 arch) return true; } +static int etm4_cpu_id(struct coresight_device *csdev) +{ + struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + return drvdata->cpu; +} + static int etm4_trace_id(struct coresight_device *csdev) { struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); @@ -262,6 +269,7 @@ static void etm4_disable(struct coresight_device *csdev) } static const struct coresight_ops_source etm4_source_ops = { + .cpu_id = etm4_cpu_id, .trace_id = etm4_trace_id, .enable = etm4_enable, .disable = etm4_disable, diff --git a/include/linux/coresight.h b/include/linux/coresight.h index a7cabfa23b55..bf62b265bf52 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -205,12 +205,15 @@ struct coresight_ops_link { /** * struct coresight_ops_source - basic operations for a source * Operations available for sources. + * @cpu_id: returns the value of the CPU number this component + * is associated to. * @trace_id: returns the value of the component's trace ID as known to the HW. * @enable: enables tracing for a source. * @disable: disables tracing for a source. */ struct coresight_ops_source { + int (*cpu_id)(struct coresight_device *csdev); int (*trace_id)(struct coresight_device *csdev); int (*enable)(struct coresight_device *csdev); void (*disable)(struct coresight_device *csdev); -- cgit From 5f74f82ea34c0da80ea0b49192bb5ea06e063593 Mon Sep 17 00:00:00 2001 From: Hans Westgaard Ry Date: Wed, 3 Feb 2016 09:26:57 +0100 Subject: net:Add sysctl_max_skb_frags MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Devices may have limits on the number of fragments in an skb they support. Current codebase uses a constant as maximum for number of fragments one skb can hold and use. When enabling scatter/gather and running traffic with many small messages the codebase uses the maximum number of fragments and may thereby violate the max for certain devices. The patch introduces a global variable as max number of fragments. Signed-off-by: Hans Westgaard Ry Reviewed-by: Håkon Bugge Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 1 + net/core/skbuff.c | 2 ++ net/core/sysctl_net_core.c | 10 ++++++++++ net/ipv4/tcp.c | 4 ++-- 4 files changed, 15 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 11f935c1a090..4ce9ff7086f4 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -299,6 +299,7 @@ struct sk_buff; #else #define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1) #endif +extern int sysctl_max_skb_frags; typedef struct skb_frag_struct skb_frag_t; diff --git a/net/core/skbuff.c b/net/core/skbuff.c index b2df375ec9c2..5bf88f58bee7 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -79,6 +79,8 @@ struct kmem_cache *skbuff_head_cache __read_mostly; static struct kmem_cache *skbuff_fclone_cache __read_mostly; +int sysctl_max_skb_frags __read_mostly = MAX_SKB_FRAGS; +EXPORT_SYMBOL(sysctl_max_skb_frags); /** * skb_panic - private function for out-of-line support diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index 95b6139d710c..a6beb7b6ae55 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -26,6 +26,7 @@ static int zero = 0; static int one = 1; static int min_sndbuf = SOCK_MIN_SNDBUF; static int min_rcvbuf = SOCK_MIN_RCVBUF; +static int max_skb_frags = MAX_SKB_FRAGS; static int net_msg_warn; /* Unused, but still a sysctl */ @@ -392,6 +393,15 @@ static struct ctl_table net_core_table[] = { .mode = 0644, .proc_handler = proc_dointvec }, + { + .procname = "max_skb_frags", + .data = &sysctl_max_skb_frags, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &one, + .extra2 = &max_skb_frags, + }, { } }; diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 19746b3fcbbe..0c36ef4a3f86 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -940,7 +940,7 @@ new_segment: i = skb_shinfo(skb)->nr_frags; can_coalesce = skb_can_coalesce(skb, i, page, offset); - if (!can_coalesce && i >= MAX_SKB_FRAGS) { + if (!can_coalesce && i >= sysctl_max_skb_frags) { tcp_mark_push(tp, skb); goto new_segment; } @@ -1213,7 +1213,7 @@ new_segment: if (!skb_can_coalesce(skb, i, pfrag->page, pfrag->offset)) { - if (i == MAX_SKB_FRAGS || !sg) { + if (i == sysctl_max_skb_frags || !sg) { tcp_mark_push(tp, skb); goto new_segment; } -- cgit From 4a389810bc3cb0e73443104f0827e81e23cb1e12 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 11 Feb 2016 16:13:14 -0800 Subject: kernel/locking/lockdep.c: convert hash tables to hlists Mike said: : CONFIG_UBSAN_ALIGNMENT breaks x86-64 kernel with lockdep enabled, i. e : kernel with CONFIG_UBSAN_ALIGNMENT fails to load without even any error : message. : : The problem is that ubsan callbacks use spinlocks and might be called : before lockdep is initialized. Particularly this line in the : reserve_ebda_region function causes problem: : : lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); : : If i put lockdep_init() before reserve_ebda_region call in : x86_64_start_reservations kernel loads well. Fix this ordering issue permanently: change lockdep so that it uses hlists for the hash tables. Unlike a list_head, an hlist_head is in its initialized state when it is all-zeroes, so lockdep is ready for operation immediately upon boot - lockdep_init() need not have run. The patch will also save some memory. lockdep_init() and lockdep_initialized can be done away with now - a 4.6 patch has been prepared to do this. Reported-by: Mike Krinkin Suggested-by: Mike Krinkin Cc: Andrey Ryabinin Cc: Ingo Molnar Cc: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/lockdep.h | 4 ++-- kernel/locking/lockdep.c | 42 +++++++++++++++++++----------------------- 2 files changed, 21 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index c57e424d914b..4dca42fd32f5 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -66,7 +66,7 @@ struct lock_class { /* * class-hash: */ - struct list_head hash_entry; + struct hlist_node hash_entry; /* * global list of all lock-classes: @@ -199,7 +199,7 @@ struct lock_chain { u8 irq_context; u8 depth; u16 base; - struct list_head entry; + struct hlist_node entry; u64 chain_key; }; diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 60ace56618f6..7537e568dabe 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -292,7 +292,7 @@ LIST_HEAD(all_lock_classes); #define __classhashfn(key) hash_long((unsigned long)key, CLASSHASH_BITS) #define classhashentry(key) (classhash_table + __classhashfn((key))) -static struct list_head classhash_table[CLASSHASH_SIZE]; +static struct hlist_head classhash_table[CLASSHASH_SIZE]; /* * We put the lock dependency chains into a hash-table as well, to cache @@ -303,7 +303,7 @@ static struct list_head classhash_table[CLASSHASH_SIZE]; #define __chainhashfn(chain) hash_long(chain, CHAINHASH_BITS) #define chainhashentry(chain) (chainhash_table + __chainhashfn((chain))) -static struct list_head chainhash_table[CHAINHASH_SIZE]; +static struct hlist_head chainhash_table[CHAINHASH_SIZE]; /* * The hash key of the lock dependency chains is a hash itself too: @@ -666,7 +666,7 @@ static inline struct lock_class * look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) { struct lockdep_subclass_key *key; - struct list_head *hash_head; + struct hlist_head *hash_head; struct lock_class *class; #ifdef CONFIG_DEBUG_LOCKDEP @@ -719,7 +719,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) return NULL; - list_for_each_entry_rcu(class, hash_head, hash_entry) { + hlist_for_each_entry_rcu(class, hash_head, hash_entry) { if (class->key == key) { /* * Huh! same key, different name? Did someone trample @@ -742,7 +742,7 @@ static inline struct lock_class * register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) { struct lockdep_subclass_key *key; - struct list_head *hash_head; + struct hlist_head *hash_head; struct lock_class *class; DEBUG_LOCKS_WARN_ON(!irqs_disabled()); @@ -774,7 +774,7 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) * We have to do the hash-walk again, to avoid races * with another CPU: */ - list_for_each_entry_rcu(class, hash_head, hash_entry) { + hlist_for_each_entry_rcu(class, hash_head, hash_entry) { if (class->key == key) goto out_unlock_set; } @@ -805,7 +805,7 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) * We use RCU's safe list-add method to make * parallel walking of the hash-list safe: */ - list_add_tail_rcu(&class->hash_entry, hash_head); + hlist_add_head_rcu(&class->hash_entry, hash_head); /* * Add it to the global list of classes: */ @@ -2017,7 +2017,7 @@ static inline int lookup_chain_cache(struct task_struct *curr, u64 chain_key) { struct lock_class *class = hlock_class(hlock); - struct list_head *hash_head = chainhashentry(chain_key); + struct hlist_head *hash_head = chainhashentry(chain_key); struct lock_chain *chain; struct held_lock *hlock_curr; int i, j; @@ -2033,7 +2033,7 @@ static inline int lookup_chain_cache(struct task_struct *curr, * We can walk it lock-free, because entries only get added * to the hash: */ - list_for_each_entry_rcu(chain, hash_head, entry) { + hlist_for_each_entry_rcu(chain, hash_head, entry) { if (chain->chain_key == chain_key) { cache_hit: debug_atomic_inc(chain_lookup_hits); @@ -2057,7 +2057,7 @@ cache_hit: /* * We have to walk the chain again locked - to avoid duplicates: */ - list_for_each_entry(chain, hash_head, entry) { + hlist_for_each_entry(chain, hash_head, entry) { if (chain->chain_key == chain_key) { graph_unlock(); goto cache_hit; @@ -2091,7 +2091,7 @@ cache_hit: } chain_hlocks[chain->base + j] = class - lock_classes; } - list_add_tail_rcu(&chain->entry, hash_head); + hlist_add_head_rcu(&chain->entry, hash_head); debug_atomic_inc(chain_lookup_misses); inc_chains(); @@ -3875,7 +3875,7 @@ void lockdep_reset(void) nr_process_chains = 0; debug_locks = 1; for (i = 0; i < CHAINHASH_SIZE; i++) - INIT_LIST_HEAD(chainhash_table + i); + INIT_HLIST_HEAD(chainhash_table + i); raw_local_irq_restore(flags); } @@ -3894,7 +3894,7 @@ static void zap_class(struct lock_class *class) /* * Unhash the class and remove it from the all_lock_classes list: */ - list_del_rcu(&class->hash_entry); + hlist_del_rcu(&class->hash_entry); list_del_rcu(&class->lock_entry); RCU_INIT_POINTER(class->key, NULL); @@ -3917,7 +3917,7 @@ static inline int within(const void *addr, void *start, unsigned long size) void lockdep_free_key_range(void *start, unsigned long size) { struct lock_class *class; - struct list_head *head; + struct hlist_head *head; unsigned long flags; int i; int locked; @@ -3930,9 +3930,7 @@ void lockdep_free_key_range(void *start, unsigned long size) */ for (i = 0; i < CLASSHASH_SIZE; i++) { head = classhash_table + i; - if (list_empty(head)) - continue; - list_for_each_entry_rcu(class, head, hash_entry) { + hlist_for_each_entry_rcu(class, head, hash_entry) { if (within(class->key, start, size)) zap_class(class); else if (within(class->name, start, size)) @@ -3962,7 +3960,7 @@ void lockdep_free_key_range(void *start, unsigned long size) void lockdep_reset_lock(struct lockdep_map *lock) { struct lock_class *class; - struct list_head *head; + struct hlist_head *head; unsigned long flags; int i, j; int locked; @@ -3987,9 +3985,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) locked = graph_lock(); for (i = 0; i < CLASSHASH_SIZE; i++) { head = classhash_table + i; - if (list_empty(head)) - continue; - list_for_each_entry_rcu(class, head, hash_entry) { + hlist_for_each_entry_rcu(class, head, hash_entry) { int match = 0; for (j = 0; j < NR_LOCKDEP_CACHING_CLASSES; j++) @@ -4027,10 +4023,10 @@ void lockdep_init(void) return; for (i = 0; i < CLASSHASH_SIZE; i++) - INIT_LIST_HEAD(classhash_table + i); + INIT_HLIST_HEAD(classhash_table + i); for (i = 0; i < CHAINHASH_SIZE; i++) - INIT_LIST_HEAD(chainhash_table + i); + INIT_HLIST_HEAD(chainhash_table + i); lockdep_initialized = 1; } -- cgit From db78c22230d0bcc8b27b81f05b39f104f08232c5 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 11 Feb 2016 16:13:17 -0800 Subject: mm: fix pfn_t vs highmem The pfn_t type uses an unsigned long to store a pfn + flags value. On a 64-bit platform the upper 12 bits of an unsigned long are never used for storing the value of a pfn. However, this is not true on highmem platforms, all 32-bits of a pfn value are used to address a 44-bit physical address space. A pfn_t needs to store a 64-bit value. Link: https://bugzilla.kernel.org/show_bug.cgi?id=112211 Fixes: 01c8f1c44b83 ("mm, dax, gpu: convert vm_insert_mixed to pfn_t") Signed-off-by: Dan Williams Reported-by: Stuart Foster Reported-by: Julian Margetson Tested-by: Julian Margetson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/pfn.h | 2 +- include/linux/pfn_t.h | 19 +++++++++---------- kernel/memremap.c | 2 +- 3 files changed, 11 insertions(+), 12 deletions(-) (limited to 'include/linux') diff --git a/include/linux/pfn.h b/include/linux/pfn.h index 2d8e49711b63..1132953235c0 100644 --- a/include/linux/pfn.h +++ b/include/linux/pfn.h @@ -10,7 +10,7 @@ * backing is indicated by flags in the high bits of the value. */ typedef struct { - unsigned long val; + u64 val; } pfn_t; #endif diff --git a/include/linux/pfn_t.h b/include/linux/pfn_t.h index 37448ab5fb5c..94994810c7c0 100644 --- a/include/linux/pfn_t.h +++ b/include/linux/pfn_t.h @@ -9,14 +9,13 @@ * PFN_DEV - pfn is not covered by system memmap by default * PFN_MAP - pfn has a dynamic page mapping established by a device driver */ -#define PFN_FLAGS_MASK (((unsigned long) ~PAGE_MASK) \ - << (BITS_PER_LONG - PAGE_SHIFT)) -#define PFN_SG_CHAIN (1UL << (BITS_PER_LONG - 1)) -#define PFN_SG_LAST (1UL << (BITS_PER_LONG - 2)) -#define PFN_DEV (1UL << (BITS_PER_LONG - 3)) -#define PFN_MAP (1UL << (BITS_PER_LONG - 4)) - -static inline pfn_t __pfn_to_pfn_t(unsigned long pfn, unsigned long flags) +#define PFN_FLAGS_MASK (((u64) ~PAGE_MASK) << (BITS_PER_LONG_LONG - PAGE_SHIFT)) +#define PFN_SG_CHAIN (1ULL << (BITS_PER_LONG_LONG - 1)) +#define PFN_SG_LAST (1ULL << (BITS_PER_LONG_LONG - 2)) +#define PFN_DEV (1ULL << (BITS_PER_LONG_LONG - 3)) +#define PFN_MAP (1ULL << (BITS_PER_LONG_LONG - 4)) + +static inline pfn_t __pfn_to_pfn_t(unsigned long pfn, u64 flags) { pfn_t pfn_t = { .val = pfn | (flags & PFN_FLAGS_MASK), }; @@ -29,7 +28,7 @@ static inline pfn_t pfn_to_pfn_t(unsigned long pfn) return __pfn_to_pfn_t(pfn, 0); } -extern pfn_t phys_to_pfn_t(phys_addr_t addr, unsigned long flags); +extern pfn_t phys_to_pfn_t(phys_addr_t addr, u64 flags); static inline bool pfn_t_has_page(pfn_t pfn) { @@ -87,7 +86,7 @@ static inline pmd_t pfn_t_pmd(pfn_t pfn, pgprot_t pgprot) #ifdef __HAVE_ARCH_PTE_DEVMAP static inline bool pfn_t_devmap(pfn_t pfn) { - const unsigned long flags = PFN_DEV|PFN_MAP; + const u64 flags = PFN_DEV|PFN_MAP; return (pfn.val & flags) == flags; } diff --git a/kernel/memremap.c b/kernel/memremap.c index 70ee3775de24..2c468dea60bc 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -150,7 +150,7 @@ void devm_memunmap(struct device *dev, void *addr) } EXPORT_SYMBOL(devm_memunmap); -pfn_t phys_to_pfn_t(phys_addr_t addr, unsigned long flags) +pfn_t phys_to_pfn_t(phys_addr_t addr, u64 flags) { return __pfn_to_pfn_t(addr >> PAGE_SHIFT, flags); } -- cgit From e1379b56e9e88653fcb58cbaa71cd6b1cc304918 Mon Sep 17 00:00:00 2001 From: Cory Tusar Date: Wed, 10 Feb 2016 14:32:07 -0500 Subject: misc: eeprom_93xx46: Add quirks to support Atmel AT93C46D device. Atmel devices in this family have some quirks not found in other similar chips - they do not support a sequential read of the entire EEPROM contents, and the control word sent at the start of each operation varies in bit length. This commit adds quirk support to the driver and modifies the read implementation to support non-sequential reads for consistency with other misc/eeprom drivers. Tested on a custom Freescale VF610-based platform, with an AT93C46D device attached via dspi2. The spi-gpio driver was used to allow the necessary non-byte-sized transfers. Signed-off-by: Cory Tusar Tested-by: Chris Healy Reviewed-by: Vladimir Zapolskiy Signed-off-by: Greg Kroah-Hartman --- drivers/misc/eeprom/eeprom_93xx46.c | 126 ++++++++++++++++++++++++++---------- include/linux/eeprom_93xx46.h | 6 ++ 2 files changed, 97 insertions(+), 35 deletions(-) (limited to 'include/linux') diff --git a/drivers/misc/eeprom/eeprom_93xx46.c b/drivers/misc/eeprom/eeprom_93xx46.c index f0c30b366218..a3c3136abf3e 100644 --- a/drivers/misc/eeprom/eeprom_93xx46.c +++ b/drivers/misc/eeprom/eeprom_93xx46.c @@ -27,6 +27,15 @@ #define ADDR_ERAL 0x20 #define ADDR_EWEN 0x30 +struct eeprom_93xx46_devtype_data { + unsigned int quirks; +}; + +static const struct eeprom_93xx46_devtype_data atmel_at93c46d_data = { + .quirks = EEPROM_93XX46_QUIRK_SINGLE_WORD_READ | + EEPROM_93XX46_QUIRK_INSTRUCTION_LENGTH, +}; + struct eeprom_93xx46_dev { struct spi_device *spi; struct eeprom_93xx46_platform_data *pdata; @@ -35,6 +44,16 @@ struct eeprom_93xx46_dev { int addrlen; }; +static inline bool has_quirk_single_word_read(struct eeprom_93xx46_dev *edev) +{ + return edev->pdata->quirks & EEPROM_93XX46_QUIRK_SINGLE_WORD_READ; +} + +static inline bool has_quirk_instruction_length(struct eeprom_93xx46_dev *edev) +{ + return edev->pdata->quirks & EEPROM_93XX46_QUIRK_INSTRUCTION_LENGTH; +} + static ssize_t eeprom_93xx46_bin_read(struct file *filp, struct kobject *kobj, struct bin_attribute *bin_attr, @@ -42,58 +61,73 @@ eeprom_93xx46_bin_read(struct file *filp, struct kobject *kobj, { struct eeprom_93xx46_dev *edev; struct device *dev; - struct spi_message m; - struct spi_transfer t[2]; - int bits, ret; - u16 cmd_addr; + ssize_t ret = 0; dev = kobj_to_dev(kobj); edev = dev_get_drvdata(dev); - cmd_addr = OP_READ << edev->addrlen; + mutex_lock(&edev->lock); - if (edev->addrlen == 7) { - cmd_addr |= off & 0x7f; - bits = 10; - } else { - cmd_addr |= (off >> 1) & 0x3f; - bits = 9; - } + if (edev->pdata->prepare) + edev->pdata->prepare(edev); - dev_dbg(&edev->spi->dev, "read cmd 0x%x, %d Hz\n", - cmd_addr, edev->spi->max_speed_hz); + while (count) { + struct spi_message m; + struct spi_transfer t[2] = { { 0 } }; + u16 cmd_addr = OP_READ << edev->addrlen; + size_t nbytes = count; + int bits; + int err; + + if (edev->addrlen == 7) { + cmd_addr |= off & 0x7f; + bits = 10; + if (has_quirk_single_word_read(edev)) + nbytes = 1; + } else { + cmd_addr |= (off >> 1) & 0x3f; + bits = 9; + if (has_quirk_single_word_read(edev)) + nbytes = 2; + } - spi_message_init(&m); - memset(t, 0, sizeof(t)); + dev_dbg(&edev->spi->dev, "read cmd 0x%x, %d Hz\n", + cmd_addr, edev->spi->max_speed_hz); - t[0].tx_buf = (char *)&cmd_addr; - t[0].len = 2; - t[0].bits_per_word = bits; - spi_message_add_tail(&t[0], &m); + spi_message_init(&m); - t[1].rx_buf = buf; - t[1].len = count; - t[1].bits_per_word = 8; - spi_message_add_tail(&t[1], &m); + t[0].tx_buf = (char *)&cmd_addr; + t[0].len = 2; + t[0].bits_per_word = bits; + spi_message_add_tail(&t[0], &m); - mutex_lock(&edev->lock); + t[1].rx_buf = buf; + t[1].len = count; + t[1].bits_per_word = 8; + spi_message_add_tail(&t[1], &m); - if (edev->pdata->prepare) - edev->pdata->prepare(edev); + err = spi_sync(edev->spi, &m); + /* have to wait at least Tcsl ns */ + ndelay(250); - ret = spi_sync(edev->spi, &m); - /* have to wait at least Tcsl ns */ - ndelay(250); - if (ret) { - dev_err(&edev->spi->dev, "read %zu bytes at %d: err. %d\n", - count, (int)off, ret); + if (err) { + dev_err(&edev->spi->dev, "read %zu bytes at %d: err. %d\n", + nbytes, (int)off, err); + ret = err; + break; + } + + buf += nbytes; + off += nbytes; + count -= nbytes; + ret += nbytes; } if (edev->pdata->finish) edev->pdata->finish(edev); mutex_unlock(&edev->lock); - return ret ? : count; + return ret; } static int eeprom_93xx46_ew(struct eeprom_93xx46_dev *edev, int is_on) @@ -112,7 +146,13 @@ static int eeprom_93xx46_ew(struct eeprom_93xx46_dev *edev, int is_on) bits = 9; } - dev_dbg(&edev->spi->dev, "ew cmd 0x%04x\n", cmd_addr); + if (has_quirk_instruction_length(edev)) { + cmd_addr <<= 2; + bits += 2; + } + + dev_dbg(&edev->spi->dev, "ew%s cmd 0x%04x, %d bits\n", + is_on ? "en" : "ds", cmd_addr, bits); spi_message_init(&m); memset(&t, 0, sizeof(t)); @@ -247,6 +287,13 @@ static int eeprom_93xx46_eral(struct eeprom_93xx46_dev *edev) bits = 9; } + if (has_quirk_instruction_length(edev)) { + cmd_addr <<= 2; + bits += 2; + } + + dev_dbg(&edev->spi->dev, "eral cmd 0x%04x, %d bits\n", cmd_addr, bits); + spi_message_init(&m); memset(&t, 0, sizeof(t)); @@ -298,12 +345,15 @@ static DEVICE_ATTR(erase, S_IWUSR, NULL, eeprom_93xx46_store_erase); static const struct of_device_id eeprom_93xx46_of_table[] = { { .compatible = "eeprom-93xx46", }, + { .compatible = "atmel,at93c46d", .data = &atmel_at93c46d_data, }, {} }; MODULE_DEVICE_TABLE(of, eeprom_93xx46_of_table); static int eeprom_93xx46_probe_dt(struct spi_device *spi) { + const struct of_device_id *of_id = + of_match_device(eeprom_93xx46_of_table, &spi->dev); struct device_node *np = spi->dev.of_node; struct eeprom_93xx46_platform_data *pd; u32 tmp; @@ -331,6 +381,12 @@ static int eeprom_93xx46_probe_dt(struct spi_device *spi) if (of_property_read_bool(np, "read-only")) pd->flags |= EE_READONLY; + if (of_id->data) { + const struct eeprom_93xx46_devtype_data *data = of_id->data; + + pd->quirks = data->quirks; + } + spi->dev.platform_data = pd; return 0; diff --git a/include/linux/eeprom_93xx46.h b/include/linux/eeprom_93xx46.h index 06791811e49d..92fa4c37ac1f 100644 --- a/include/linux/eeprom_93xx46.h +++ b/include/linux/eeprom_93xx46.h @@ -9,6 +9,12 @@ struct eeprom_93xx46_platform_data { #define EE_ADDR16 0x02 /* 16 bit addr. cfg */ #define EE_READONLY 0x08 /* forbid writing */ + unsigned int quirks; +/* Single word read transfers only; no sequential read. */ +#define EEPROM_93XX46_QUIRK_SINGLE_WORD_READ (1 << 0) +/* Instructions such as EWEN are (addrlen + 2) in length. */ +#define EEPROM_93XX46_QUIRK_INSTRUCTION_LENGTH (1 << 1) + /* * optional hooks to control additional logic * before and after spi transfer. -- cgit From 3ca9b1ac28398c6fe0bed335d2d71a35e1c5f7c9 Mon Sep 17 00:00:00 2001 From: Cory Tusar Date: Wed, 10 Feb 2016 14:32:08 -0500 Subject: misc: eeprom_93xx46: Add support for a GPIO 'select' line. This commit adds support to the eeprom_93x46 driver allowing a GPIO line to function as a 'select' or 'enable' signal prior to accessing the EEPROM. Signed-off-by: Cory Tusar Tested-by: Chris Healy Reviewed-by: Vladimir Zapolskiy Signed-off-by: Greg Kroah-Hartman --- drivers/misc/eeprom/eeprom_93xx46.c | 35 +++++++++++++++++++++++++++++++++++ include/linux/eeprom_93xx46.h | 3 +++ 2 files changed, 38 insertions(+) (limited to 'include/linux') diff --git a/drivers/misc/eeprom/eeprom_93xx46.c b/drivers/misc/eeprom/eeprom_93xx46.c index a3c3136abf3e..f62ab29e293c 100644 --- a/drivers/misc/eeprom/eeprom_93xx46.c +++ b/drivers/misc/eeprom/eeprom_93xx46.c @@ -10,11 +10,13 @@ #include #include +#include #include #include #include #include #include +#include #include #include #include @@ -343,6 +345,20 @@ static ssize_t eeprom_93xx46_store_erase(struct device *dev, } static DEVICE_ATTR(erase, S_IWUSR, NULL, eeprom_93xx46_store_erase); +static void select_assert(void *context) +{ + struct eeprom_93xx46_dev *edev = context; + + gpiod_set_value_cansleep(edev->pdata->select, 1); +} + +static void select_deassert(void *context) +{ + struct eeprom_93xx46_dev *edev = context; + + gpiod_set_value_cansleep(edev->pdata->select, 0); +} + static const struct of_device_id eeprom_93xx46_of_table[] = { { .compatible = "eeprom-93xx46", }, { .compatible = "atmel,at93c46d", .data = &atmel_at93c46d_data, }, @@ -357,6 +373,8 @@ static int eeprom_93xx46_probe_dt(struct spi_device *spi) struct device_node *np = spi->dev.of_node; struct eeprom_93xx46_platform_data *pd; u32 tmp; + int gpio; + enum of_gpio_flags of_flags; int ret; pd = devm_kzalloc(&spi->dev, sizeof(*pd), GFP_KERNEL); @@ -381,6 +399,23 @@ static int eeprom_93xx46_probe_dt(struct spi_device *spi) if (of_property_read_bool(np, "read-only")) pd->flags |= EE_READONLY; + gpio = of_get_named_gpio_flags(np, "select-gpios", 0, &of_flags); + if (gpio_is_valid(gpio)) { + unsigned long flags = + of_flags == OF_GPIO_ACTIVE_LOW ? GPIOF_ACTIVE_LOW : 0; + + ret = devm_gpio_request_one(&spi->dev, gpio, flags, + "eeprom_93xx46_select"); + if (ret) + return ret; + + pd->select = gpio_to_desc(gpio); + pd->prepare = select_assert; + pd->finish = select_deassert; + + gpiod_direction_output(pd->select, 0); + } + if (of_id->data) { const struct eeprom_93xx46_devtype_data *data = of_id->data; diff --git a/include/linux/eeprom_93xx46.h b/include/linux/eeprom_93xx46.h index 92fa4c37ac1f..885f587a3555 100644 --- a/include/linux/eeprom_93xx46.h +++ b/include/linux/eeprom_93xx46.h @@ -3,6 +3,8 @@ * platform description for 93xx46 EEPROMs. */ +struct gpio_desc; + struct eeprom_93xx46_platform_data { unsigned char flags; #define EE_ADDR8 0x01 /* 8 bit addr. cfg */ @@ -21,4 +23,5 @@ struct eeprom_93xx46_platform_data { */ void (*prepare)(void *); void (*finish)(void *); + struct gpio_desc *select; }; -- cgit From f8560a9bc76b2cd5c06fa412cb7b5481d70fcf34 Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Mon, 15 Feb 2016 19:12:01 +0200 Subject: stm class: Use driver's packet callback return value STM drivers provide a callback to generate/send individual STP packets; it also tells the stm core how many bytes of payload it has consumed. However, we would also need to use the negative space of this return value to communicate errors that occur during the packet generation, in which case the stm core will have to take appropriate action. For now, we need to account for the possibility that the stm driver may not support certain combinations of packet type/flags, in which case it is expected to signal an error. Signed-off-by: Alexander Shishkin Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/stm/core.c | 19 ++++++++++++------- include/linux/stm.h | 7 +++++++ 2 files changed, 19 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/stm/core.c b/drivers/hwtracing/stm/core.c index 79cca94bfb58..0db303b50e51 100644 --- a/drivers/hwtracing/stm/core.c +++ b/drivers/hwtracing/stm/core.c @@ -380,8 +380,8 @@ static int stm_file_assign(struct stm_file *stmf, char *id, unsigned int width) return ret; } -static void stm_write(struct stm_data *data, unsigned int master, - unsigned int channel, const char *buf, size_t count) +static ssize_t stm_write(struct stm_data *data, unsigned int master, + unsigned int channel, const char *buf, size_t count) { unsigned int flags = STP_PACKET_TIMESTAMPED; const unsigned char *p = buf, nil = 0; @@ -393,9 +393,14 @@ static void stm_write(struct stm_data *data, unsigned int master, sz = data->packet(data, master, channel, STP_PACKET_DATA, flags, sz, p); flags = 0; + + if (sz < 0) + break; } data->packet(data, master, channel, STP_PACKET_FLAG, 0, 0, &nil); + + return pos; } static ssize_t stm_char_write(struct file *file, const char __user *buf, @@ -433,8 +438,8 @@ static ssize_t stm_char_write(struct file *file, const char __user *buf, return -EFAULT; } - stm_write(stm->data, stmf->output.master, stmf->output.channel, kbuf, - count); + count = stm_write(stm->data, stmf->output.master, stmf->output.channel, + kbuf, count); kfree(kbuf); @@ -996,9 +1001,9 @@ int stm_source_write(struct stm_source_data *data, unsigned int chan, stm = srcu_dereference(src->link, &stm_source_srcu); if (stm) - stm_write(stm->data, src->output.master, - src->output.channel + chan, - buf, count); + count = stm_write(stm->data, src->output.master, + src->output.channel + chan, + buf, count); else count = -ENODEV; diff --git a/include/linux/stm.h b/include/linux/stm.h index 9d0083d364e6..ab8ceca4f570 100644 --- a/include/linux/stm.h +++ b/include/linux/stm.h @@ -67,6 +67,13 @@ struct stm_device; * description. That is, the lowest master that can be allocated to software * writers is @sw_start and data from this writer will appear is @sw_start * master in the STP stream. + * + * The @packet callback should adhere to the following rules: + * 1) it must return the number of bytes it consumed from the payload; + * 2) therefore, if it sent a packet that does not have payload (like FLAG), + * it must return zero; + * 3) if it does not support the requested packet type/flag combination, + * it must return -ENOTSUPP. */ struct stm_data { const char *name; -- cgit From cc8424074e51355e0c6ba717d8edc50d408f2802 Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Mon, 15 Feb 2016 19:12:09 +0200 Subject: stm class: Plug stm device's unlink callback STM device's unlink callback is never actually called from anywhere in the stm class code. This patch adds calls to stm driver's unlink method after the unlinking has succeeded. Signed-off-by: Alexander Shishkin Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/stm/core.c | 23 +++++++++++++++++++---- include/linux/stm.h | 3 +++ 2 files changed, 22 insertions(+), 4 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/stm/core.c b/drivers/hwtracing/stm/core.c index 30181821d909..de80d45d8df9 100644 --- a/drivers/hwtracing/stm/core.c +++ b/drivers/hwtracing/stm/core.c @@ -376,14 +376,19 @@ err_free: static int stm_char_release(struct inode *inode, struct file *file) { struct stm_file *stmf = file->private_data; + struct stm_device *stm = stmf->stm; + + if (stm->data->unlink) + stm->data->unlink(stm->data, stmf->output.master, + stmf->output.channel); - stm_output_free(stmf->stm, &stmf->output); + stm_output_free(stm, &stmf->output); /* * matches the stm_char_open()'s * class_find_device() + try_module_get() */ - stm_put_device(stmf->stm); + stm_put_device(stm); kfree(stmf); return 0; @@ -865,8 +870,18 @@ unlock: spin_unlock(&src->link_lock); spin_unlock(&stm->link_lock); - if (!ret && src->data->unlink) - src->data->unlink(src->data); + /* + * Call the unlink callbacks for both source and stm, when we know + * that we have actually performed the unlinking. + */ + if (!ret) { + if (src->data->unlink) + src->data->unlink(src->data); + + if (stm->data->unlink) + stm->data->unlink(stm->data, src->output.master, + src->output.channel); + } return ret; } diff --git a/include/linux/stm.h b/include/linux/stm.h index ab8ceca4f570..1a79ed8e43da 100644 --- a/include/linux/stm.h +++ b/include/linux/stm.h @@ -74,6 +74,9 @@ struct stm_device; * it must return zero; * 3) if it does not support the requested packet type/flag combination, * it must return -ENOTSUPP. + * + * The @unlink callback is called when there are no more active writers so + * that the master/channel can be quiesced. */ struct stm_data { const char *name; -- cgit From b3e94405941e6916d5e365454d74560c2bea47ca Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:51:45 -0700 Subject: coresight: associating path with session rather than tracer When using the Coresight framework from the sysFS interface a tracer is always handling a single session and as such, a path can be associated with a tracer. But when supporting multiple session per tracer there is no guarantee that sessions will always have the same path from source to sink. This patch is removing the automatic association between path and tracers. The building of a path and enablement of the components in the path are decoupled, allowing for the association of a path with a session rather than a tracer. To keep backward functionality with the current sysFS access methods a per-cpu place holder is used to keep a handle on the path built when tracers are enabled. Lastly APIs to build paths and enable tracers are made public so that other subsystem can interact with the Coresight framework. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-priv.h | 5 + drivers/hwtracing/coresight/coresight.c | 296 ++++++++++++++++++--------- include/linux/coresight.h | 2 - 3 files changed, 206 insertions(+), 97 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h index 62fcd98cc7cf..7b193a34d709 100644 --- a/drivers/hwtracing/coresight/coresight-priv.h +++ b/drivers/hwtracing/coresight/coresight-priv.h @@ -52,6 +52,11 @@ static inline void CS_UNLOCK(void __iomem *addr) } while (0); } +void coresight_disable_path(struct list_head *path); +int coresight_enable_path(struct list_head *path); +struct list_head *coresight_build_path(struct coresight_device *csdev); +void coresight_release_path(struct list_head *path); + #ifdef CONFIG_CORESIGHT_SOURCE_ETM3X extern int etm_readl_cp14(u32 off, unsigned int *val); extern int etm_writel_cp14(u32 off, u32 val); diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index 7e6e9ff27dd1..f26589effb70 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -29,6 +29,22 @@ static DEFINE_MUTEX(coresight_mutex); +/** + * struct coresight_node - elements of a path, from source to sink + * @csdev: Address of an element. + * @link: hook to the list. + */ +struct coresight_node { + struct coresight_device *csdev; + struct list_head link; +}; + +/* + * When operating Coresight drivers from the sysFS interface, only a single + * path can exist from a tracer (associated to a CPU) to a sink. + */ +static DEFINE_PER_CPU(struct list_head *, sysfs_path); + static int coresight_id_match(struct device *dev, void *data) { int trace_id, i_trace_id; @@ -68,15 +84,12 @@ static int coresight_source_is_unique(struct coresight_device *csdev) csdev, coresight_id_match); } -static int coresight_find_link_inport(struct coresight_device *csdev) +static int coresight_find_link_inport(struct coresight_device *csdev, + struct coresight_device *parent) { int i; - struct coresight_device *parent; struct coresight_connection *conn; - parent = container_of(csdev->path_link.next, - struct coresight_device, path_link); - for (i = 0; i < parent->nr_outport; i++) { conn = &parent->conns[i]; if (conn->child_dev == csdev) @@ -89,15 +102,12 @@ static int coresight_find_link_inport(struct coresight_device *csdev) return 0; } -static int coresight_find_link_outport(struct coresight_device *csdev) +static int coresight_find_link_outport(struct coresight_device *csdev, + struct coresight_device *child) { int i; - struct coresight_device *child; struct coresight_connection *conn; - child = container_of(csdev->path_link.prev, - struct coresight_device, path_link); - for (i = 0; i < csdev->nr_outport; i++) { conn = &csdev->conns[i]; if (conn->child_dev == child) @@ -138,14 +148,19 @@ static void coresight_disable_sink(struct coresight_device *csdev) } } -static int coresight_enable_link(struct coresight_device *csdev) +static int coresight_enable_link(struct coresight_device *csdev, + struct coresight_device *parent, + struct coresight_device *child) { int ret; int link_subtype; int refport, inport, outport; - inport = coresight_find_link_inport(csdev); - outport = coresight_find_link_outport(csdev); + if (!parent || !child) + return -EINVAL; + + inport = coresight_find_link_inport(csdev, parent); + outport = coresight_find_link_outport(csdev, child); link_subtype = csdev->subtype.link_subtype; if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG) @@ -168,14 +183,19 @@ static int coresight_enable_link(struct coresight_device *csdev) return 0; } -static void coresight_disable_link(struct coresight_device *csdev) +static void coresight_disable_link(struct coresight_device *csdev, + struct coresight_device *parent, + struct coresight_device *child) { int i, nr_conns; int link_subtype; int refport, inport, outport; - inport = coresight_find_link_inport(csdev); - outport = coresight_find_link_outport(csdev); + if (!parent || !child) + return; + + inport = coresight_find_link_inport(csdev, parent); + outport = coresight_find_link_outport(csdev, child); link_subtype = csdev->subtype.link_subtype; if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG) { @@ -235,109 +255,167 @@ static void coresight_disable_source(struct coresight_device *csdev) } } -static int coresight_enable_path(struct list_head *path) +void coresight_disable_path(struct list_head *path) { - int ret = 0; - struct coresight_device *cd; + struct coresight_node *nd; + struct coresight_device *csdev, *parent, *child; - /* - * At this point we have a full @path, from source to sink. The - * sink is the first entry and the source the last one. Go through - * all the components and enable them one by one. - */ - list_for_each_entry(cd, path, path_link) { - if (cd == list_first_entry(path, struct coresight_device, - path_link)) { - ret = coresight_enable_sink(cd); - } else if (list_is_last(&cd->path_link, path)) { - /* - * Don't enable the source just yet - this needs to - * happen at the very end when all links and sink - * along the path have been configured properly. - */ - ; - } else { - ret = coresight_enable_link(cd); + list_for_each_entry(nd, path, link) { + csdev = nd->csdev; + + switch (csdev->type) { + case CORESIGHT_DEV_TYPE_SINK: + case CORESIGHT_DEV_TYPE_LINKSINK: + coresight_disable_sink(csdev); + break; + case CORESIGHT_DEV_TYPE_SOURCE: + /* sources are disabled from either sysFS or Perf */ + break; + case CORESIGHT_DEV_TYPE_LINK: + parent = list_prev_entry(nd, link)->csdev; + child = list_next_entry(nd, link)->csdev; + coresight_disable_link(csdev, parent, child); + break; + default: + break; } - if (ret) - goto err; } +} - return 0; -err: - list_for_each_entry_continue_reverse(cd, path, path_link) { - if (cd == list_first_entry(path, struct coresight_device, - path_link)) { - coresight_disable_sink(cd); - } else if (list_is_last(&cd->path_link, path)) { - ; - } else { - coresight_disable_link(cd); +int coresight_enable_path(struct list_head *path) +{ + + int ret = 0; + struct coresight_node *nd; + struct coresight_device *csdev, *parent, *child; + + list_for_each_entry_reverse(nd, path, link) { + csdev = nd->csdev; + + switch (csdev->type) { + case CORESIGHT_DEV_TYPE_SINK: + case CORESIGHT_DEV_TYPE_LINKSINK: + ret = coresight_enable_sink(csdev); + if (ret) + goto err; + break; + case CORESIGHT_DEV_TYPE_SOURCE: + /* sources are enabled from either sysFS or Perf */ + break; + case CORESIGHT_DEV_TYPE_LINK: + parent = list_prev_entry(nd, link)->csdev; + child = list_next_entry(nd, link)->csdev; + ret = coresight_enable_link(csdev, parent, child); + if (ret) + goto err; + break; + default: + goto err; } } +out: return ret; +err: + coresight_disable_path(path); + goto out; } -static int coresight_disable_path(struct list_head *path) +/** + * _coresight_build_path - recursively build a path from a @csdev to a sink. + * @csdev: The device to start from. + * @path: The list to add devices to. + * + * The tree of Coresight device is traversed until an activated sink is + * found. From there the sink is added to the list along with all the + * devices that led to that point - the end result is a list from source + * to sink. In that list the source is the first device and the sink the + * last one. + */ +static int _coresight_build_path(struct coresight_device *csdev, + struct list_head *path) { - struct coresight_device *cd; + int i; + bool found = false; + struct coresight_node *node; + struct coresight_connection *conn; - list_for_each_entry_reverse(cd, path, path_link) { - if (cd == list_first_entry(path, struct coresight_device, - path_link)) { - coresight_disable_sink(cd); - } else if (list_is_last(&cd->path_link, path)) { - /* - * The source has already been stopped, no need - * to do it again here. - */ - ; - } else { - coresight_disable_link(cd); + /* An activated sink has been found. Enqueue the element */ + if ((csdev->type == CORESIGHT_DEV_TYPE_SINK || + csdev->type == CORESIGHT_DEV_TYPE_LINKSINK) && csdev->activated) + goto out; + + /* Not a sink - recursively explore each port found on this element */ + for (i = 0; i < csdev->nr_outport; i++) { + conn = &csdev->conns[i]; + if (_coresight_build_path(conn->child_dev, path) == 0) { + found = true; + break; } } + if (!found) + return -ENODEV; + +out: + /* + * A path from this element to a sink has been found. The elements + * leading to the sink are already enqueued, all that is left to do + * is add a node for this element. + */ + node = kzalloc(sizeof(struct coresight_node), GFP_KERNEL); + if (!node) + return -ENOMEM; + + node->csdev = csdev; + list_add(&node->link, path); + return 0; } -static int coresight_build_paths(struct coresight_device *csdev, - struct list_head *path, - bool enable) +struct list_head *coresight_build_path(struct coresight_device *csdev) { - int i, ret = -EINVAL; - struct coresight_connection *conn; + struct list_head *path; - list_add(&csdev->path_link, path); + path = kzalloc(sizeof(struct list_head), GFP_KERNEL); + if (!path) + return NULL; - if ((csdev->type == CORESIGHT_DEV_TYPE_SINK || - csdev->type == CORESIGHT_DEV_TYPE_LINKSINK) && - csdev->activated) { - if (enable) - ret = coresight_enable_path(path); - else - ret = coresight_disable_path(path); - } else { - for (i = 0; i < csdev->nr_outport; i++) { - conn = &csdev->conns[i]; - if (coresight_build_paths(conn->child_dev, - path, enable) == 0) - ret = 0; - } + INIT_LIST_HEAD(path); + + if (_coresight_build_path(csdev, path)) { + kfree(path); + path = NULL; } - if (list_first_entry(path, struct coresight_device, path_link) != csdev) - dev_err(&csdev->dev, "wrong device in %s\n", __func__); + return path; +} - list_del(&csdev->path_link); +/** + * coresight_release_path - release a previously built path. + * @path: the path to release. + * + * Go through all the elements of a path and 1) removed it from the list and + * 2) free the memory allocated for each node. + */ +void coresight_release_path(struct list_head *path) +{ + struct coresight_node *nd, *next; - return ret; + list_for_each_entry_safe(nd, next, path, link) { + list_del(&nd->link); + kfree(nd); + } + + kfree(path); + path = NULL; } int coresight_enable(struct coresight_device *csdev) { int ret = 0; - LIST_HEAD(path); + int cpu; + struct list_head *path; mutex_lock(&coresight_mutex); if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) { @@ -348,22 +426,47 @@ int coresight_enable(struct coresight_device *csdev) if (csdev->enable) goto out; - if (coresight_build_paths(csdev, &path, true)) { - dev_err(&csdev->dev, "building path(s) failed\n"); + path = coresight_build_path(csdev); + if (!path) { + pr_err("building path(s) failed\n"); goto out; } - if (coresight_enable_source(csdev)) - dev_err(&csdev->dev, "source enable failed\n"); + ret = coresight_enable_path(path); + if (ret) + goto err_path; + + ret = coresight_enable_source(csdev); + if (ret) + goto err_source; + + /* + * When working from sysFS it is important to keep track + * of the paths that were created so that they can be + * undone in 'coresight_disable()'. Since there can only + * be a single session per tracer (when working from sysFS) + * a per-cpu variable will do just fine. + */ + cpu = source_ops(csdev)->cpu_id(csdev); + per_cpu(sysfs_path, cpu) = path; + out: mutex_unlock(&coresight_mutex); return ret; + +err_source: + coresight_disable_path(path); + +err_path: + coresight_release_path(path); + goto out; } EXPORT_SYMBOL_GPL(coresight_enable); void coresight_disable(struct coresight_device *csdev) { - LIST_HEAD(path); + int cpu; + struct list_head *path; mutex_lock(&coresight_mutex); if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) { @@ -373,9 +476,12 @@ void coresight_disable(struct coresight_device *csdev) if (!csdev->enable) goto out; + cpu = source_ops(csdev)->cpu_id(csdev); + path = per_cpu(sysfs_path, cpu); coresight_disable_source(csdev); - if (coresight_build_paths(csdev, &path, false)) - dev_err(&csdev->dev, "releasing path(s) failed\n"); + coresight_disable_path(path); + coresight_release_path(path); + per_cpu(sysfs_path, cpu) = NULL; out: mutex_unlock(&coresight_mutex); diff --git a/include/linux/coresight.h b/include/linux/coresight.h index bf62b265bf52..851ecb22397e 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -152,7 +152,6 @@ struct coresight_connection { by @coresight_ops. * @dev: The device entity associated to this component. * @refcnt: keep track of what is in use. - * @path_link: link of current component into the path being enabled. * @orphan: true if the component has connections that haven't been linked. * @enable: 'true' if component is currently part of an active path. * @activated: 'true' only if a _sink_ has been activated. A sink can be @@ -168,7 +167,6 @@ struct coresight_device { const struct coresight_ops *ops; struct device dev; atomic_t *refcnt; - struct list_head path_link; bool orphan; bool enable; /* true only if configured as part of a path */ bool activated; /* true only if a sink is part of a path */ -- cgit From 22fd532eaa0c24d86e23d8e9e3b7feac4a8cac80 Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:51:52 -0700 Subject: coresight: etm3x: adding operation mode for etm_enable() Adding a new mode to source API enable() in order to distinguish where the request comes from. That way it is possible to perform different operations based on where the request was issued from. The ETM4x driver is also modified to keep in sync with the new interface. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etm.h | 5 +- .../hwtracing/coresight/coresight-etm3x-sysfs.c | 4 +- drivers/hwtracing/coresight/coresight-etm3x.c | 68 +++++++++++++++++++--- drivers/hwtracing/coresight/coresight-etm4x.c | 2 +- drivers/hwtracing/coresight/coresight-priv.h | 6 ++ drivers/hwtracing/coresight/coresight.c | 6 +- include/linux/coresight.h | 2 +- 7 files changed, 76 insertions(+), 17 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etm.h b/drivers/hwtracing/coresight/coresight-etm.h index 371fb7d2e829..5b29d5540fe5 100644 --- a/drivers/hwtracing/coresight/coresight-etm.h +++ b/drivers/hwtracing/coresight/coresight-etm.h @@ -13,6 +13,7 @@ #ifndef _CORESIGHT_CORESIGHT_ETM_H #define _CORESIGHT_CORESIGHT_ETM_H +#include #include #include "coresight-priv.h" @@ -214,7 +215,7 @@ struct etm_config { * @port_size: port size as reported by ETMCR bit 4-6 and 21. * @arch: ETM/PTM version number. * @use_cpu14: true if management registers need to be accessed via CP14. - * @enable: is this ETM/PTM currently tracing. + * @mode: this tracer's mode, i.e sysFS, Perf or disabled. * @sticky_enable: true if ETM base configuration has been done. * @boot_enable:true if we should start tracing at boot time. * @os_unlock: true if access to management registers is allowed. @@ -238,7 +239,7 @@ struct etm_drvdata { int port_size; u8 arch; bool use_cp14; - bool enable; + local_t mode; bool sticky_enable; bool boot_enable; bool os_unlock; diff --git a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c index 456df2378a6f..387c79fd9d5e 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c +++ b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c @@ -716,7 +716,7 @@ static ssize_t cntr_val_show(struct device *dev, struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent); struct etm_config *config = &drvdata->config; - if (!drvdata->enable) { + if (!local_read(&drvdata->mode)) { spin_lock(&drvdata->spinlock); for (i = 0; i < drvdata->nr_cntr; i++) ret += sprintf(buf, "counter %d: %x\n", @@ -935,7 +935,7 @@ static ssize_t seq_curr_state_show(struct device *dev, struct etm_drvdata *drvdata = dev_get_drvdata(dev->parent); struct etm_config *config = &drvdata->config; - if (!drvdata->enable) { + if (!local_read(&drvdata->mode)) { val = config->seq_curr_state; goto out; } diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index 1952dc31fee4..e16501b41ed8 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -306,7 +306,7 @@ int etm_get_trace_id(struct etm_drvdata *drvdata) if (!drvdata) goto out; - if (!drvdata->enable) + if (!local_read(&drvdata->mode)) return drvdata->traceid; pm_runtime_get_sync(drvdata->dev); @@ -332,7 +332,7 @@ static int etm_trace_id(struct coresight_device *csdev) return etm_get_trace_id(drvdata); } -static int etm_enable(struct coresight_device *csdev) +static int etm_enable_sysfs(struct coresight_device *csdev) { struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); int ret; @@ -351,18 +351,44 @@ static int etm_enable(struct coresight_device *csdev) goto err; } - drvdata->enable = true; drvdata->sticky_enable = true; - spin_unlock(&drvdata->spinlock); dev_info(drvdata->dev, "ETM tracing enabled\n"); return 0; + err: spin_unlock(&drvdata->spinlock); return ret; } +static int etm_enable(struct coresight_device *csdev, u32 mode) +{ + int ret; + u32 val; + struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + val = local_cmpxchg(&drvdata->mode, CS_MODE_DISABLED, mode); + + /* Someone is already using the tracer */ + if (val) + return -EBUSY; + + switch (mode) { + case CS_MODE_SYSFS: + ret = etm_enable_sysfs(csdev); + break; + default: + ret = -EINVAL; + } + + /* The tracer didn't start */ + if (ret) + local_set(&drvdata->mode, CS_MODE_DISABLED); + + return ret; +} + static void etm_disable_hw(void *info) { int i; @@ -387,7 +413,7 @@ static void etm_disable_hw(void *info) dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu); } -static void etm_disable(struct coresight_device *csdev) +static void etm_disable_sysfs(struct coresight_device *csdev) { struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); @@ -405,7 +431,6 @@ static void etm_disable(struct coresight_device *csdev) * ensures that register writes occur when cpu is powered. */ smp_call_function_single(drvdata->cpu, etm_disable_hw, drvdata, 1); - drvdata->enable = false; spin_unlock(&drvdata->spinlock); put_online_cpus(); @@ -413,6 +438,33 @@ static void etm_disable(struct coresight_device *csdev) dev_info(drvdata->dev, "ETM tracing disabled\n"); } +static void etm_disable(struct coresight_device *csdev) +{ + u32 mode; + struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + /* + * For as long as the tracer isn't disabled another entity can't + * change its status. As such we can read the status here without + * fearing it will change under us. + */ + mode = local_read(&drvdata->mode); + + switch (mode) { + case CS_MODE_DISABLED: + break; + case CS_MODE_SYSFS: + etm_disable_sysfs(csdev); + break; + default: + WARN_ON_ONCE(mode); + return; + } + + if (mode) + local_set(&drvdata->mode, CS_MODE_DISABLED); +} + static const struct coresight_ops_source etm_source_ops = { .cpu_id = etm_cpu_id, .trace_id = etm_trace_id, @@ -440,7 +492,7 @@ static int etm_cpu_callback(struct notifier_block *nfb, unsigned long action, etmdrvdata[cpu]->os_unlock = true; } - if (etmdrvdata[cpu]->enable) + if (local_read(&etmdrvdata[cpu]->mode)) etm_enable_hw(etmdrvdata[cpu]); spin_unlock(&etmdrvdata[cpu]->spinlock); break; @@ -453,7 +505,7 @@ static int etm_cpu_callback(struct notifier_block *nfb, unsigned long action, case CPU_DYING: spin_lock(&etmdrvdata[cpu]->spinlock); - if (etmdrvdata[cpu]->enable) + if (local_read(&etmdrvdata[cpu]->mode)) etm_disable_hw(etmdrvdata[cpu]); spin_unlock(&etmdrvdata[cpu]->spinlock); break; diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c index c2f518fbc9a8..0026092fec7f 100644 --- a/drivers/hwtracing/coresight/coresight-etm4x.c +++ b/drivers/hwtracing/coresight/coresight-etm4x.c @@ -187,7 +187,7 @@ static void etm4_enable_hw(void *info) dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu); } -static int etm4_enable(struct coresight_device *csdev) +static int etm4_enable(struct coresight_device *csdev, u32 mode) { struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); int ret; diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h index 14f245a2018d..ed116b303e87 100644 --- a/drivers/hwtracing/coresight/coresight-priv.h +++ b/drivers/hwtracing/coresight/coresight-priv.h @@ -34,6 +34,12 @@ #define TIMEOUT_US 100 #define BMVAL(val, lsb, msb) ((val & GENMASK(msb, lsb)) >> lsb) +enum cs_mode { + CS_MODE_DISABLED, + CS_MODE_SYSFS, + CS_MODE_PERF, +}; + static inline void CS_LOCK(void __iomem *addr) { do { diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index 6b44928c1076..b20afb709141 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -222,7 +222,7 @@ static void coresight_disable_link(struct coresight_device *csdev, csdev->enable = false; } -static int coresight_enable_source(struct coresight_device *csdev) +static int coresight_enable_source(struct coresight_device *csdev, u32 mode) { int ret; @@ -234,7 +234,7 @@ static int coresight_enable_source(struct coresight_device *csdev) if (!csdev->enable) { if (source_ops(csdev)->enable) { - ret = source_ops(csdev)->enable(csdev); + ret = source_ops(csdev)->enable(csdev, mode); if (ret) return ret; } @@ -458,7 +458,7 @@ int coresight_enable(struct coresight_device *csdev) if (ret) goto err_path; - ret = coresight_enable_source(csdev); + ret = coresight_enable_source(csdev, CS_MODE_SYSFS); if (ret) goto err_source; diff --git a/include/linux/coresight.h b/include/linux/coresight.h index 851ecb22397e..61dfb8d511ea 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -213,7 +213,7 @@ struct coresight_ops_link { struct coresight_ops_source { int (*cpu_id)(struct coresight_device *csdev); int (*trace_id)(struct coresight_device *csdev); - int (*enable)(struct coresight_device *csdev); + int (*enable)(struct coresight_device *csdev, u32 mode); void (*disable)(struct coresight_device *csdev); }; -- cgit From 882d5e112491c875ab7c8c336b8beaeec54d0509 Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:51:57 -0700 Subject: coresight: etm3x: implementing perf_enable/disable() API That way traces can be enabled and disabled automatically from the Perf subystem using the PMU abstraction. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/Kconfig | 1 + drivers/hwtracing/coresight/coresight-etm3x.c | 95 +++++++++++++++++++++++++-- drivers/hwtracing/coresight/coresight-etm4x.c | 4 +- drivers/hwtracing/coresight/coresight.c | 2 +- include/linux/coresight.h | 6 +- 5 files changed, 99 insertions(+), 9 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/Kconfig b/drivers/hwtracing/coresight/Kconfig index c85935f3525a..db0541031c72 100644 --- a/drivers/hwtracing/coresight/Kconfig +++ b/drivers/hwtracing/coresight/Kconfig @@ -4,6 +4,7 @@ menuconfig CORESIGHT bool "CoreSight Tracing Support" select ARM_AMBA + select PERF_EVENTS help This framework provides a kernel interface for the CoreSight debug and trace drivers to register themselves with. It's intended to build diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index c82d545e68ef..a9b820ec16aa 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include "coresight-etm.h" @@ -297,6 +298,47 @@ void etm_config_trace_mode(struct etm_config *config) config->addr_type[1] = ETM_ADDR_TYPE_RANGE; } +#define ETM3X_SUPPORTED_OPTIONS (ETMCR_CYC_ACC | ETMCR_TIMESTAMP_EN) + +static int etm_parse_event_config(struct etm_drvdata *drvdata, + struct perf_event_attr *attr) +{ + struct etm_config *config = &drvdata->config; + + if (!attr) + return -EINVAL; + + /* Clear configuration from previous run */ + memset(config, 0, sizeof(struct etm_config)); + + if (attr->exclude_kernel) + config->mode = ETM_MODE_EXCL_KERN; + + if (attr->exclude_user) + config->mode = ETM_MODE_EXCL_USER; + + /* Always start from the default config */ + etm_set_default(config); + + /* + * By default the tracers are configured to trace the whole address + * range. Narrow the field only if requested by user space. + */ + if (config->mode) + etm_config_trace_mode(config); + + /* + * At this time only cycle accurate and timestamp options are + * available. + */ + if (attr->config & ~ETM3X_SUPPORTED_OPTIONS) + return -EINVAL; + + config->ctrl = attr->config; + + return 0; +} + static void etm_enable_hw(void *info) { int i; @@ -316,8 +358,10 @@ static void etm_enable_hw(void *info) etm_set_prog(drvdata); etmcr = etm_readl(drvdata, ETMCR); - etmcr &= (ETMCR_PWD_DWN | ETMCR_ETM_PRG); + /* Clear setting from a previous run if need be */ + etmcr &= ~ETM3X_SUPPORTED_OPTIONS; etmcr |= drvdata->port_size; + etmcr |= ETMCR_ETM_EN; etm_writel(drvdata, config->ctrl | etmcr, ETMCR); etm_writel(drvdata, config->trigger_event, ETMTRIGGER); etm_writel(drvdata, config->startstop_ctrl, ETMTSSCR); @@ -357,9 +401,6 @@ static void etm_enable_hw(void *info) /* No VMID comparator value selected */ etm_writel(drvdata, 0x0, ETMVMIDCVR); - /* Ensures trace output is enabled from this ETM */ - etm_writel(drvdata, config->ctrl | ETMCR_ETM_EN | etmcr, ETMCR); - etm_clr_prog(drvdata); CS_LOCK(drvdata->base); @@ -407,6 +448,22 @@ static int etm_trace_id(struct coresight_device *csdev) return etm_get_trace_id(drvdata); } +static int etm_enable_perf(struct coresight_device *csdev, + struct perf_event_attr *attr) +{ + struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id())) + return -EINVAL; + + /* Configure the tracer based on the session's specifics */ + etm_parse_event_config(drvdata, attr); + /* And enable it */ + etm_enable_hw(drvdata); + + return 0; +} + static int etm_enable_sysfs(struct coresight_device *csdev) { struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); @@ -437,7 +494,8 @@ err: return ret; } -static int etm_enable(struct coresight_device *csdev, u32 mode) +static int etm_enable(struct coresight_device *csdev, + struct perf_event_attr *attr, u32 mode) { int ret; u32 val; @@ -453,6 +511,9 @@ static int etm_enable(struct coresight_device *csdev, u32 mode) case CS_MODE_SYSFS: ret = etm_enable_sysfs(csdev); break; + case CS_MODE_PERF: + ret = etm_enable_perf(csdev, attr); + break; default: ret = -EINVAL; } @@ -485,6 +546,27 @@ static void etm_disable_hw(void *info) dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu); } +static void etm_disable_perf(struct coresight_device *csdev) +{ + struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id())) + return; + + CS_UNLOCK(drvdata->base); + + /* Setting the prog bit disables tracing immediately */ + etm_set_prog(drvdata); + + /* + * There is no way to know when the tracer will be used again so + * power down the tracer. + */ + etm_set_pwrdwn(drvdata); + + CS_LOCK(drvdata->base); +} + static void etm_disable_sysfs(struct coresight_device *csdev) { struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); @@ -528,6 +610,9 @@ static void etm_disable(struct coresight_device *csdev) case CS_MODE_SYSFS: etm_disable_sysfs(csdev); break; + case CS_MODE_PERF: + etm_disable_perf(csdev); + break; default: WARN_ON_ONCE(mode); return; diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c index 0026092fec7f..d0169ba7fbf2 100644 --- a/drivers/hwtracing/coresight/coresight-etm4x.c +++ b/drivers/hwtracing/coresight/coresight-etm4x.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include "coresight-etm4x.h" @@ -187,7 +188,8 @@ static void etm4_enable_hw(void *info) dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu); } -static int etm4_enable(struct coresight_device *csdev, u32 mode) +static int etm4_enable(struct coresight_device *csdev, + struct perf_event_attr *attr, u32 mode) { struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); int ret; diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index b20afb709141..95cccb179763 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -234,7 +234,7 @@ static int coresight_enable_source(struct coresight_device *csdev, u32 mode) if (!csdev->enable) { if (source_ops(csdev)->enable) { - ret = source_ops(csdev)->enable(csdev, mode); + ret = source_ops(csdev)->enable(csdev, NULL, mode); if (ret) return ret; } diff --git a/include/linux/coresight.h b/include/linux/coresight.h index 61dfb8d511ea..6801dd64ee5d 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -14,6 +14,7 @@ #define _LINUX_CORESIGHT_H #include +#include #include /* Peripheral id registers (0xFD0-0xFEC) */ @@ -206,14 +207,15 @@ struct coresight_ops_link { * @cpu_id: returns the value of the CPU number this component * is associated to. * @trace_id: returns the value of the component's trace ID as known - to the HW. + * to the HW. * @enable: enables tracing for a source. * @disable: disables tracing for a source. */ struct coresight_ops_source { int (*cpu_id)(struct coresight_device *csdev); int (*trace_id)(struct coresight_device *csdev); - int (*enable)(struct coresight_device *csdev, u32 mode); + int (*enable)(struct coresight_device *csdev, + struct perf_event_attr *attr, u32 mode); void (*disable)(struct coresight_device *csdev); }; -- cgit From e827d4550aa3225b8965ce4c266208cfe0297509 Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:51:59 -0700 Subject: coresight: etb10: adding operation mode for sink->enable() Adding an operation mode to the sink->enable() API in order to prevent simultaneous access from different callers. TPIU and TMC won't be supplemented with the AUX area API immediately and as such ignore the new mode. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etb10.c | 30 ++++++++++++++++++++------- drivers/hwtracing/coresight/coresight-priv.h | 2 +- drivers/hwtracing/coresight/coresight-tmc.c | 2 +- drivers/hwtracing/coresight/coresight-tpiu.c | 2 +- drivers/hwtracing/coresight/coresight.c | 10 ++++----- include/linux/coresight.h | 2 +- 6 files changed, 32 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c index 162c9ccc8c33..79099f95ba3f 100644 --- a/drivers/hwtracing/coresight/coresight-etb10.c +++ b/drivers/hwtracing/coresight/coresight-etb10.c @@ -73,9 +73,9 @@ * @miscdev: specifics to handle "/dev/xyz.etb" entry. * @spinlock: only one at a time pls. * @reading: synchronise user space access to etb buffer. + * @mode: this ETB is being used. * @buf: area of memory where ETB buffer content gets sent. * @buffer_depth: size of @buf. - * @enable: this ETB is being used. * @trigger_cntr: amount of words to store after a trigger. */ struct etb_drvdata { @@ -86,9 +86,9 @@ struct etb_drvdata { struct miscdevice miscdev; spinlock_t spinlock; local_t reading; + local_t mode; u8 *buf; u32 buffer_depth; - bool enable; u32 trigger_cntr; }; @@ -133,16 +133,31 @@ static void etb_enable_hw(struct etb_drvdata *drvdata) CS_LOCK(drvdata->base); } -static int etb_enable(struct coresight_device *csdev) +static int etb_enable(struct coresight_device *csdev, u32 mode) { - struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + u32 val; unsigned long flags; + struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + val = local_cmpxchg(&drvdata->mode, + CS_MODE_DISABLED, mode); + /* + * When accessing from Perf, a HW buffer can be handled + * by a single trace entity. In sysFS mode many tracers + * can be logging to the same HW buffer. + */ + if (val == CS_MODE_PERF) + return -EBUSY; + + /* Nothing to do, the tracer is already enabled. */ + if (val == CS_MODE_SYSFS) + goto out; spin_lock_irqsave(&drvdata->spinlock, flags); etb_enable_hw(drvdata); - drvdata->enable = true; spin_unlock_irqrestore(&drvdata->spinlock, flags); +out: dev_info(drvdata->dev, "ETB enabled\n"); return 0; } @@ -243,9 +258,10 @@ static void etb_disable(struct coresight_device *csdev) spin_lock_irqsave(&drvdata->spinlock, flags); etb_disable_hw(drvdata); etb_dump_hw(drvdata); - drvdata->enable = false; spin_unlock_irqrestore(&drvdata->spinlock, flags); + local_set(&drvdata->mode, CS_MODE_DISABLED); + dev_info(drvdata->dev, "ETB disabled\n"); } @@ -263,7 +279,7 @@ static void etb_dump(struct etb_drvdata *drvdata) unsigned long flags; spin_lock_irqsave(&drvdata->spinlock, flags); - if (drvdata->enable) { + if (local_read(&drvdata->mode) == CS_MODE_SYSFS) { etb_disable_hw(drvdata); etb_dump_hw(drvdata); etb_enable_hw(drvdata); diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h index 932f34a84d96..333eddaed339 100644 --- a/drivers/hwtracing/coresight/coresight-priv.h +++ b/drivers/hwtracing/coresight/coresight-priv.h @@ -62,7 +62,7 @@ static inline void CS_UNLOCK(void __iomem *addr) } void coresight_disable_path(struct list_head *path); -int coresight_enable_path(struct list_head *path); +int coresight_enable_path(struct list_head *path, u32 mode); struct coresight_device *coresight_get_sink(struct list_head *path); struct list_head *coresight_build_path(struct coresight_device *csdev); void coresight_release_path(struct list_head *path); diff --git a/drivers/hwtracing/coresight/coresight-tmc.c b/drivers/hwtracing/coresight/coresight-tmc.c index b526396d80b6..ac91b0b40ec2 100644 --- a/drivers/hwtracing/coresight/coresight-tmc.c +++ b/drivers/hwtracing/coresight/coresight-tmc.c @@ -265,7 +265,7 @@ static int tmc_enable(struct tmc_drvdata *drvdata, enum tmc_mode mode) return 0; } -static int tmc_enable_sink(struct coresight_device *csdev) +static int tmc_enable_sink(struct coresight_device *csdev, u32 mode) { struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); diff --git a/drivers/hwtracing/coresight/coresight-tpiu.c b/drivers/hwtracing/coresight/coresight-tpiu.c index 0d8fce651ff7..71582f6dfd4c 100644 --- a/drivers/hwtracing/coresight/coresight-tpiu.c +++ b/drivers/hwtracing/coresight/coresight-tpiu.c @@ -70,7 +70,7 @@ static void tpiu_enable_hw(struct tpiu_drvdata *drvdata) CS_LOCK(drvdata->base); } -static int tpiu_enable(struct coresight_device *csdev) +static int tpiu_enable(struct coresight_device *csdev, u32 mode) { struct tpiu_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index 95cccb179763..6ec2b66af9ee 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -121,13 +121,13 @@ static int coresight_find_link_outport(struct coresight_device *csdev, return 0; } -static int coresight_enable_sink(struct coresight_device *csdev) +static int coresight_enable_sink(struct coresight_device *csdev, u32 mode) { int ret; if (!csdev->enable) { if (sink_ops(csdev)->enable) { - ret = sink_ops(csdev)->enable(csdev); + ret = sink_ops(csdev)->enable(csdev, mode); if (ret) return ret; } @@ -283,7 +283,7 @@ void coresight_disable_path(struct list_head *path) } } -int coresight_enable_path(struct list_head *path) +int coresight_enable_path(struct list_head *path, u32 mode) { int ret = 0; @@ -296,7 +296,7 @@ int coresight_enable_path(struct list_head *path) switch (csdev->type) { case CORESIGHT_DEV_TYPE_SINK: case CORESIGHT_DEV_TYPE_LINKSINK: - ret = coresight_enable_sink(csdev); + ret = coresight_enable_sink(csdev, mode); if (ret) goto err; break; @@ -454,7 +454,7 @@ int coresight_enable(struct coresight_device *csdev) goto out; } - ret = coresight_enable_path(path); + ret = coresight_enable_path(path, CS_MODE_SYSFS); if (ret) goto err_path; diff --git a/include/linux/coresight.h b/include/linux/coresight.h index 6801dd64ee5d..9fa92dcdd2ea 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -186,7 +186,7 @@ struct coresight_device { * @disable: disables the sink. */ struct coresight_ops_sink { - int (*enable)(struct coresight_device *csdev); + int (*enable)(struct coresight_device *csdev, u32 mode); void (*disable)(struct coresight_device *csdev); }; -- cgit From 2997aa4063d97fdb39450c6078bd81a7b0504f22 Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:52:00 -0700 Subject: coresight: etb10: implementing AUX API Adding an ETB10 specific AUX area operations to be used by the perf framework when events are initialised. Part of this operation involves modeling the mmap'ed area based on the specific ways a sink buffer gathers information. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etb10.c | 234 ++++++++++++++++++++++++++ include/linux/coresight.h | 21 ++- 2 files changed, 253 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c index 79099f95ba3f..a2eb6bdeaafa 100644 --- a/drivers/hwtracing/coresight/coresight-etb10.c +++ b/drivers/hwtracing/coresight/coresight-etb10.c @@ -28,6 +28,11 @@ #include #include #include +#include +#include +#include + +#include #include "coresight-priv.h" @@ -64,6 +69,26 @@ #define ETB_FFSR_BIT 1 #define ETB_FRAME_SIZE_WORDS 4 +/** + * struct cs_buffer - keep track of a recording session' specifics + * @cur: index of the current buffer + * @nr_pages: max number of pages granted to us + * @offset: offset within the current buffer + * @data_size: how much we collected in this run + * @lost: other than zero if we had a HW buffer wrap around + * @snapshot: is this run in snapshot mode + * @data_pages: a handle the ring buffer + */ +struct cs_buffers { + unsigned int cur; + unsigned int nr_pages; + unsigned long offset; + local_t data_size; + local_t lost; + bool snapshot; + void **data_pages; +}; + /** * struct etb_drvdata - specifics associated to an ETB component * @base: memory mapped base address for this component. @@ -265,9 +290,218 @@ static void etb_disable(struct coresight_device *csdev) dev_info(drvdata->dev, "ETB disabled\n"); } +static void *etb_alloc_buffer(struct coresight_device *csdev, int cpu, + void **pages, int nr_pages, bool overwrite) +{ + int node; + struct cs_buffers *buf; + + if (cpu == -1) + cpu = smp_processor_id(); + node = cpu_to_node(cpu); + + buf = kzalloc_node(sizeof(struct cs_buffers), GFP_KERNEL, node); + if (!buf) + return NULL; + + buf->snapshot = overwrite; + buf->nr_pages = nr_pages; + buf->data_pages = pages; + + return buf; +} + +static void etb_free_buffer(void *config) +{ + struct cs_buffers *buf = config; + + kfree(buf); +} + +static int etb_set_buffer(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config) +{ + int ret = 0; + unsigned long head; + struct cs_buffers *buf = sink_config; + + /* wrap head around to the amount of space we have */ + head = handle->head & ((buf->nr_pages << PAGE_SHIFT) - 1); + + /* find the page to write to */ + buf->cur = head / PAGE_SIZE; + + /* and offset within that page */ + buf->offset = head % PAGE_SIZE; + + local_set(&buf->data_size, 0); + + return ret; +} + +static unsigned long etb_reset_buffer(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config, bool *lost) +{ + unsigned long size = 0; + struct cs_buffers *buf = sink_config; + + if (buf) { + /* + * In snapshot mode ->data_size holds the new address of the + * ring buffer's head. The size itself is the whole address + * range since we want the latest information. + */ + if (buf->snapshot) + handle->head = local_xchg(&buf->data_size, + buf->nr_pages << PAGE_SHIFT); + + /* + * Tell the tracer PMU how much we got in this run and if + * something went wrong along the way. Nobody else can use + * this cs_buffers instance until we are done. As such + * resetting parameters here and squaring off with the ring + * buffer API in the tracer PMU is fine. + */ + *lost = !!local_xchg(&buf->lost, 0); + size = local_xchg(&buf->data_size, 0); + } + + return size; +} + +static void etb_update_buffer(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config) +{ + int i, cur; + u8 *buf_ptr; + u32 read_ptr, write_ptr, capacity; + u32 status, read_data, to_read; + unsigned long offset; + struct cs_buffers *buf = sink_config; + struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent); + + if (!buf) + return; + + capacity = drvdata->buffer_depth * ETB_FRAME_SIZE_WORDS; + + CS_UNLOCK(drvdata->base); + etb_disable_hw(drvdata); + + /* unit is in words, not bytes */ + read_ptr = readl_relaxed(drvdata->base + ETB_RAM_READ_POINTER); + write_ptr = readl_relaxed(drvdata->base + ETB_RAM_WRITE_POINTER); + + /* + * Entries should be aligned to the frame size. If they are not + * go back to the last alignement point to give decoding tools a + * chance to fix things. + */ + if (write_ptr % ETB_FRAME_SIZE_WORDS) { + dev_err(drvdata->dev, + "write_ptr: %lu not aligned to formatter frame size\n", + (unsigned long)write_ptr); + + write_ptr &= ~(ETB_FRAME_SIZE_WORDS - 1); + local_inc(&buf->lost); + } + + /* + * Get a hold of the status register and see if a wrap around + * has occurred. If so adjust things accordingly. Otherwise + * start at the beginning and go until the write pointer has + * been reached. + */ + status = readl_relaxed(drvdata->base + ETB_STATUS_REG); + if (status & ETB_STATUS_RAM_FULL) { + local_inc(&buf->lost); + to_read = capacity; + read_ptr = write_ptr; + } else { + to_read = CIRC_CNT(write_ptr, read_ptr, drvdata->buffer_depth); + to_read *= ETB_FRAME_SIZE_WORDS; + } + + /* + * Make sure we don't overwrite data that hasn't been consumed yet. + * It is entirely possible that the HW buffer has more data than the + * ring buffer can currently handle. If so adjust the start address + * to take only the last traces. + * + * In snapshot mode we are looking to get the latest traces only and as + * such, we don't care about not overwriting data that hasn't been + * processed by user space. + */ + if (!buf->snapshot && to_read > handle->size) { + u32 mask = ~(ETB_FRAME_SIZE_WORDS - 1); + + /* The new read pointer must be frame size aligned */ + to_read -= handle->size & mask; + /* + * Move the RAM read pointer up, keeping in mind that + * everything is in frame size units. + */ + read_ptr = (write_ptr + drvdata->buffer_depth) - + to_read / ETB_FRAME_SIZE_WORDS; + /* Wrap around if need be*/ + read_ptr &= ~(drvdata->buffer_depth - 1); + /* let the decoder know we've skipped ahead */ + local_inc(&buf->lost); + } + + /* finally tell HW where we want to start reading from */ + writel_relaxed(read_ptr, drvdata->base + ETB_RAM_READ_POINTER); + + cur = buf->cur; + offset = buf->offset; + for (i = 0; i < to_read; i += 4) { + buf_ptr = buf->data_pages[cur] + offset; + read_data = readl_relaxed(drvdata->base + + ETB_RAM_READ_DATA_REG); + *buf_ptr++ = read_data >> 0; + *buf_ptr++ = read_data >> 8; + *buf_ptr++ = read_data >> 16; + *buf_ptr++ = read_data >> 24; + + offset += 4; + if (offset >= PAGE_SIZE) { + offset = 0; + cur++; + /* wrap around at the end of the buffer */ + cur &= buf->nr_pages - 1; + } + } + + /* reset ETB buffer for next run */ + writel_relaxed(0x0, drvdata->base + ETB_RAM_READ_POINTER); + writel_relaxed(0x0, drvdata->base + ETB_RAM_WRITE_POINTER); + + /* + * In snapshot mode all we have to do is communicate to + * perf_aux_output_end() the address of the current head. In full + * trace mode the same function expects a size to move rb->aux_head + * forward. + */ + if (buf->snapshot) + local_set(&buf->data_size, (cur * PAGE_SIZE) + offset); + else + local_add(to_read, &buf->data_size); + + etb_enable_hw(drvdata); + CS_LOCK(drvdata->base); +} + static const struct coresight_ops_sink etb_sink_ops = { .enable = etb_enable, .disable = etb_disable, + .alloc_buffer = etb_alloc_buffer, + .free_buffer = etb_free_buffer, + .set_buffer = etb_set_buffer, + .reset_buffer = etb_reset_buffer, + .update_buffer = etb_update_buffer, }; static const struct coresight_ops etb_cs_ops = { diff --git a/include/linux/coresight.h b/include/linux/coresight.h index 9fa92dcdd2ea..385d62e64abb 100644 --- a/include/linux/coresight.h +++ b/include/linux/coresight.h @@ -182,12 +182,29 @@ struct coresight_device { /** * struct coresight_ops_sink - basic operations for a sink * Operations available for sinks - * @enable: enables the sink. - * @disable: disables the sink. + * @enable: enables the sink. + * @disable: disables the sink. + * @alloc_buffer: initialises perf's ring buffer for trace collection. + * @free_buffer: release memory allocated in @get_config. + * @set_buffer: initialises buffer mechanic before a trace session. + * @reset_buffer: finalises buffer mechanic after a trace session. + * @update_buffer: update buffer pointers after a trace session. */ struct coresight_ops_sink { int (*enable)(struct coresight_device *csdev, u32 mode); void (*disable)(struct coresight_device *csdev); + void *(*alloc_buffer)(struct coresight_device *csdev, int cpu, + void **pages, int nr_pages, bool overwrite); + void (*free_buffer)(void *config); + int (*set_buffer)(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config); + unsigned long (*reset_buffer)(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config, bool *lost); + void (*update_buffer)(struct coresight_device *csdev, + struct perf_output_handle *handle, + void *sink_config); }; /** -- cgit From 0bcbf2e30ff2271b54f54c8697a185f7d86ec6e4 Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:52:01 -0700 Subject: coresight: etm-perf: new PMU driver for ETM tracers Perf is a well known and used tool for performance monitoring and much more. A such it is an ideal candidate for integration with coresight based HW tracing. This patch introduces a PMU that represent a coresight tracer to the Perf core. Cc: Alexander Shishkin Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/Makefile | 3 +- drivers/hwtracing/coresight/coresight-etm-perf.c | 393 +++++++++++++++++++++++ drivers/hwtracing/coresight/coresight-etm-perf.h | 32 ++ drivers/hwtracing/coresight/coresight-etm3x.c | 7 + include/linux/coresight-pmu.h | 27 ++ 5 files changed, 461 insertions(+), 1 deletion(-) create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.c create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.h create mode 100644 include/linux/coresight-pmu.h (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/Makefile b/drivers/hwtracing/coresight/Makefile index 233d66cf22d3..cf8c6d689747 100644 --- a/drivers/hwtracing/coresight/Makefile +++ b/drivers/hwtracing/coresight/Makefile @@ -9,6 +9,7 @@ obj-$(CONFIG_CORESIGHT_SINK_ETBV10) += coresight-etb10.o obj-$(CONFIG_CORESIGHT_LINKS_AND_SINKS) += coresight-funnel.o \ coresight-replicator.o obj-$(CONFIG_CORESIGHT_SOURCE_ETM3X) += coresight-etm3x.o coresight-etm-cp14.o \ - coresight-etm3x-sysfs.o + coresight-etm3x-sysfs.o \ + coresight-etm-perf.o obj-$(CONFIG_CORESIGHT_SOURCE_ETM4X) += coresight-etm4x.o obj-$(CONFIG_CORESIGHT_QCOM_REPLICATOR) += coresight-replicator-qcom.o diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.c b/drivers/hwtracing/coresight/coresight-etm-perf.c new file mode 100644 index 000000000000..36153a77e982 --- /dev/null +++ b/drivers/hwtracing/coresight/coresight-etm-perf.c @@ -0,0 +1,393 @@ +/* + * Copyright(C) 2015 Linaro Limited. All rights reserved. + * Author: Mathieu Poirier + * + * 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. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see . + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "coresight-priv.h" + +static struct pmu etm_pmu; +static bool etm_perf_up; + +/** + * struct etm_event_data - Coresight specifics associated to an event + * @work: Handle to free allocated memory outside IRQ context. + * @mask: Hold the CPU(s) this event was set for. + * @snk_config: The sink configuration. + * @path: An array of path, each slot for one CPU. + */ +struct etm_event_data { + struct work_struct work; + cpumask_t mask; + void *snk_config; + struct list_head **path; +}; + +static DEFINE_PER_CPU(struct perf_output_handle, ctx_handle); +static DEFINE_PER_CPU(struct coresight_device *, csdev_src); + +/* ETMv3.5/PTM's ETMCR is 'config' */ +PMU_FORMAT_ATTR(cycacc, "config:" __stringify(ETM_OPT_CYCACC)); +PMU_FORMAT_ATTR(timestamp, "config:" __stringify(ETM_OPT_TS)); + +static struct attribute *etm_config_formats_attr[] = { + &format_attr_cycacc.attr, + &format_attr_timestamp.attr, + NULL, +}; + +static struct attribute_group etm_pmu_format_group = { + .name = "format", + .attrs = etm_config_formats_attr, +}; + +static const struct attribute_group *etm_pmu_attr_groups[] = { + &etm_pmu_format_group, + NULL, +}; + +static void etm_event_read(struct perf_event *event) {} + +static int etm_event_init(struct perf_event *event) +{ + if (event->attr.type != etm_pmu.type) + return -ENOENT; + + return 0; +} + +static void free_event_data(struct work_struct *work) +{ + int cpu; + cpumask_t *mask; + struct etm_event_data *event_data; + struct coresight_device *sink; + + event_data = container_of(work, struct etm_event_data, work); + mask = &event_data->mask; + /* + * First deal with the sink configuration. See comment in + * etm_setup_aux() about why we take the first available path. + */ + if (event_data->snk_config) { + cpu = cpumask_first(mask); + sink = coresight_get_sink(event_data->path[cpu]); + if (sink_ops(sink)->free_buffer) + sink_ops(sink)->free_buffer(event_data->snk_config); + } + + for_each_cpu(cpu, mask) { + if (event_data->path[cpu]) + coresight_release_path(event_data->path[cpu]); + } + + kfree(event_data->path); + kfree(event_data); +} + +static void *alloc_event_data(int cpu) +{ + int size; + cpumask_t *mask; + struct etm_event_data *event_data; + + /* First get memory for the session's data */ + event_data = kzalloc(sizeof(struct etm_event_data), GFP_KERNEL); + if (!event_data) + return NULL; + + /* Make sure nothing disappears under us */ + get_online_cpus(); + size = num_online_cpus(); + + mask = &event_data->mask; + if (cpu != -1) + cpumask_set_cpu(cpu, mask); + else + cpumask_copy(mask, cpu_online_mask); + put_online_cpus(); + + /* + * Each CPU has a single path between source and destination. As such + * allocate an array using CPU numbers as indexes. That way a path + * for any CPU can easily be accessed at any given time. We proceed + * the same way for sessions involving a single CPU. The cost of + * unused memory when dealing with single CPU trace scenarios is small + * compared to the cost of searching through an optimized array. + */ + event_data->path = kcalloc(size, + sizeof(struct list_head *), GFP_KERNEL); + if (!event_data->path) { + kfree(event_data); + return NULL; + } + + return event_data; +} + +static void etm_free_aux(void *data) +{ + struct etm_event_data *event_data = data; + + schedule_work(&event_data->work); +} + +static void *etm_setup_aux(int event_cpu, void **pages, + int nr_pages, bool overwrite) +{ + int cpu; + cpumask_t *mask; + struct coresight_device *sink; + struct etm_event_data *event_data = NULL; + + event_data = alloc_event_data(event_cpu); + if (!event_data) + return NULL; + + INIT_WORK(&event_data->work, free_event_data); + + mask = &event_data->mask; + + /* Setup the path for each CPU in a trace session */ + for_each_cpu(cpu, mask) { + struct coresight_device *csdev; + + csdev = per_cpu(csdev_src, cpu); + if (!csdev) + goto err; + + /* + * Building a path doesn't enable it, it simply builds a + * list of devices from source to sink that can be + * referenced later when the path is actually needed. + */ + event_data->path[cpu] = coresight_build_path(csdev); + if (!event_data->path[cpu]) + goto err; + } + + /* + * In theory nothing prevent tracers in a trace session from being + * associated with different sinks, nor having a sink per tracer. But + * until we have HW with this kind of topology and a way to convey + * sink assignement from the perf cmd line we need to assume tracers + * in a trace session are using the same sink. Therefore pick the sink + * found at the end of the first available path. + */ + cpu = cpumask_first(mask); + /* Grab the sink at the end of the path */ + sink = coresight_get_sink(event_data->path[cpu]); + if (!sink) + goto err; + + if (!sink_ops(sink)->alloc_buffer) + goto err; + + /* Get the AUX specific data from the sink buffer */ + event_data->snk_config = + sink_ops(sink)->alloc_buffer(sink, cpu, pages, + nr_pages, overwrite); + if (!event_data->snk_config) + goto err; + +out: + return event_data; + +err: + etm_free_aux(event_data); + event_data = NULL; + goto out; +} + +static void etm_event_start(struct perf_event *event, int flags) +{ + int cpu = smp_processor_id(); + struct etm_event_data *event_data; + struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle); + struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu); + + if (!csdev) + goto fail; + + /* + * Deal with the ring buffer API and get a handle on the + * session's information. + */ + event_data = perf_aux_output_begin(handle, event); + if (!event_data) + goto fail; + + /* We need a sink, no need to continue without one */ + sink = coresight_get_sink(event_data->path[cpu]); + if (WARN_ON_ONCE(!sink || !sink_ops(sink)->set_buffer)) + goto fail_end_stop; + + /* Configure the sink */ + if (sink_ops(sink)->set_buffer(sink, handle, + event_data->snk_config)) + goto fail_end_stop; + + /* Nothing will happen without a path */ + if (coresight_enable_path(event_data->path[cpu], CS_MODE_PERF)) + goto fail_end_stop; + + /* Tell the perf core the event is alive */ + event->hw.state = 0; + + /* Finally enable the tracer */ + if (source_ops(csdev)->enable(csdev, &event->attr, CS_MODE_PERF)) + goto fail_end_stop; + +out: + return; + +fail_end_stop: + perf_aux_output_end(handle, 0, true); +fail: + event->hw.state = PERF_HES_STOPPED; + goto out; +} + +static void etm_event_stop(struct perf_event *event, int mode) +{ + bool lost; + int cpu = smp_processor_id(); + unsigned long size; + struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu); + struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle); + struct etm_event_data *event_data = perf_get_aux(handle); + + if (event->hw.state == PERF_HES_STOPPED) + return; + + if (!csdev) + return; + + sink = coresight_get_sink(event_data->path[cpu]); + if (!sink) + return; + + /* stop tracer */ + source_ops(csdev)->disable(csdev); + + /* tell the core */ + event->hw.state = PERF_HES_STOPPED; + + if (mode & PERF_EF_UPDATE) { + if (WARN_ON_ONCE(handle->event != event)) + return; + + /* update trace information */ + if (!sink_ops(sink)->update_buffer) + return; + + sink_ops(sink)->update_buffer(sink, handle, + event_data->snk_config); + + if (!sink_ops(sink)->reset_buffer) + return; + + size = sink_ops(sink)->reset_buffer(sink, handle, + event_data->snk_config, + &lost); + + perf_aux_output_end(handle, size, lost); + } + + /* Disabling the path make its elements available to other sessions */ + coresight_disable_path(event_data->path[cpu]); +} + +static int etm_event_add(struct perf_event *event, int mode) +{ + int ret = 0; + struct hw_perf_event *hwc = &event->hw; + + if (mode & PERF_EF_START) { + etm_event_start(event, 0); + if (hwc->state & PERF_HES_STOPPED) + ret = -EINVAL; + } else { + hwc->state = PERF_HES_STOPPED; + } + + return ret; +} + +static void etm_event_del(struct perf_event *event, int mode) +{ + etm_event_stop(event, PERF_EF_UPDATE); +} + +int etm_perf_symlink(struct coresight_device *csdev, bool link) +{ + char entry[sizeof("cpu9999999")]; + int ret = 0, cpu = source_ops(csdev)->cpu_id(csdev); + struct device *pmu_dev = etm_pmu.dev; + struct device *cs_dev = &csdev->dev; + + sprintf(entry, "cpu%d", cpu); + + if (!etm_perf_up) + return -EPROBE_DEFER; + + if (link) { + ret = sysfs_create_link(&pmu_dev->kobj, &cs_dev->kobj, entry); + if (ret) + return ret; + per_cpu(csdev_src, cpu) = csdev; + } else { + sysfs_remove_link(&pmu_dev->kobj, entry); + per_cpu(csdev_src, cpu) = NULL; + } + + return 0; +} + +static int __init etm_perf_init(void) +{ + int ret; + + etm_pmu.capabilities = PERF_PMU_CAP_EXCLUSIVE; + + etm_pmu.attr_groups = etm_pmu_attr_groups; + etm_pmu.task_ctx_nr = perf_sw_context; + etm_pmu.read = etm_event_read; + etm_pmu.event_init = etm_event_init; + etm_pmu.setup_aux = etm_setup_aux; + etm_pmu.free_aux = etm_free_aux; + etm_pmu.start = etm_event_start; + etm_pmu.stop = etm_event_stop; + etm_pmu.add = etm_event_add; + etm_pmu.del = etm_event_del; + + ret = perf_pmu_register(&etm_pmu, CORESIGHT_ETM_PMU_NAME, -1); + if (ret == 0) + etm_perf_up = true; + + return ret; +} +module_init(etm_perf_init); diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.h b/drivers/hwtracing/coresight/coresight-etm-perf.h new file mode 100644 index 000000000000..87f5a134eb6f --- /dev/null +++ b/drivers/hwtracing/coresight/coresight-etm-perf.h @@ -0,0 +1,32 @@ +/* + * Copyright(C) 2015 Linaro Limited. All rights reserved. + * Author: Mathieu Poirier + * + * 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. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see . + */ + +#ifndef _CORESIGHT_ETM_PERF_H +#define _CORESIGHT_ETM_PERF_H + +struct coresight_device; + +#ifdef CONFIG_CORESIGHT +int etm_perf_symlink(struct coresight_device *csdev, bool link); + +#else +static inline int etm_perf_symlink(struct coresight_device *csdev, bool link) +{ return -EINVAL; } + +#endif /* CONFIG_CORESIGHT */ + +#endif diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index a9b820ec16aa..77b37413803f 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -35,6 +35,7 @@ #include #include "coresight-etm.h" +#include "coresight-etm-perf.h" static int boot_enable; module_param_named(boot_enable, boot_enable, int, S_IRUGO); @@ -827,6 +828,12 @@ static int etm_probe(struct amba_device *adev, const struct amba_id *id) goto err_arch_supported; } + ret = etm_perf_symlink(drvdata->csdev, true); + if (ret) { + coresight_unregister(drvdata->csdev); + goto err_arch_supported; + } + pm_runtime_put(&adev->dev); dev_info(dev, "%s initialized\n", (char *)id->data); diff --git a/include/linux/coresight-pmu.h b/include/linux/coresight-pmu.h new file mode 100644 index 000000000000..6c5386b23b10 --- /dev/null +++ b/include/linux/coresight-pmu.h @@ -0,0 +1,27 @@ +/* + * Copyright(C) 2015 Linaro Limited. All rights reserved. + * Author: Mathieu Poirier + * + * 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. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see . + */ + +#ifndef _LINUX_CORESIGHT_PMU_H +#define _LINUX_CORESIGHT_PMU_H + +#define CORESIGHT_ETM_PMU_NAME "cs_etm" + +/* ETMv3.5/PTM's ETMCR config bit */ +#define ETM_OPT_CYCACC 12 +#define ETM_OPT_TS 28 + +#endif -- cgit From 17534ceb835a1a96eb921a2a80df168723d6570a Mon Sep 17 00:00:00 2001 From: Mathieu Poirier Date: Wed, 17 Feb 2016 17:52:02 -0700 Subject: coresight: introducing a global trace ID function TraceID values have to be unique for all tracers and consistent between drivers and user space. As such introducing a central function to be used whenever a traceID value is required. The patch also account for data traceIDs, which are usually I(N) + 1. Signed-off-by: Mathieu Poirier Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etm3x.c | 7 ++----- include/linux/coresight-pmu.h | 12 ++++++++++++ 2 files changed, 14 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index 77b37413803f..0ba1a3981373 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include #include @@ -740,11 +741,7 @@ static void etm_init_arch_data(void *info) static void etm_init_trace_id(struct etm_drvdata *drvdata) { - /* - * A trace ID of value 0 is invalid, so let's start at some - * random value that fits in 7 bits and go from there. - */ - drvdata->traceid = 0x10 + drvdata->cpu; + drvdata->traceid = coresight_get_trace_id(drvdata->cpu); } static int etm_probe(struct amba_device *adev, const struct amba_id *id) diff --git a/include/linux/coresight-pmu.h b/include/linux/coresight-pmu.h index 6c5386b23b10..7d410260661b 100644 --- a/include/linux/coresight-pmu.h +++ b/include/linux/coresight-pmu.h @@ -19,9 +19,21 @@ #define _LINUX_CORESIGHT_PMU_H #define CORESIGHT_ETM_PMU_NAME "cs_etm" +#define CORESIGHT_ETM_PMU_SEED 0x10 /* ETMv3.5/PTM's ETMCR config bit */ #define ETM_OPT_CYCACC 12 #define ETM_OPT_TS 28 +static inline int coresight_get_trace_id(int cpu) +{ + /* + * A trace ID of value 0 is invalid, so let's start at some + * random value that fits in 7 bits and go from there. Since + * the common convention is to have data trace IDs be I(N) + 1, + * set instruction trace IDs as a function of the CPU number. + */ + return (CORESIGHT_ETM_PMU_SEED + (cpu * 2)); +} + #endif -- cgit From 941943cf519f7cacbbcecee5c4ef4b77b466bd5c Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Wed, 17 Feb 2016 17:52:03 -0700 Subject: drivers/hwtracing: make coresight-* explicitly non-modular None of the Kconfig currently controlling compilation of any of the files here are tristate, meaning that none of it currently is being built as a module by anyone. We need not be concerned about .remove functions and blocking the unbind sysfs operations, since that was already done in a recent commit. Lets remove any remaining modular references, so that when reading the drivers there is no doubt they are builtin-only. All drivers get mostly the same changes, so they are handled in batch. Changes are (1) convert to builtin_amba_driver, (2) delete module.h include where unused, and (3) relocate the description into the comments so we don't need MODULE_DESCRIPTION and associated tags. The etm3x and etm4x use module_param_named, and have been adjusted to just include moduleparam.h for that purpose. In commit f309d4443130bf814e991f836e919dca22df37ae ("platform_device: better support builtin boilerplate avoidance") we introduced the builtin_driver macro. Here we use that support and extend it to amba driver registration, so where a driver is clearly non-modular and builtin-only, we can update with the simple mapping of module_amba_driver(...) ---> builtin_amba_driver(...) Since module_amba_driver() uses the same init level priority as builtin_amba_driver() the init ordering remains unchanged with this commit. Cc: Mathieu Poirier Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Paul Gortmaker Signed-off-by: Mathieu Poirier Acked-by: Russell King Signed-off-by: Greg Kroah-Hartman --- drivers/hwtracing/coresight/coresight-etb10.c | 9 +++------ drivers/hwtracing/coresight/coresight-etm3x.c | 14 ++++++++------ drivers/hwtracing/coresight/coresight-etm4x.c | 4 +--- drivers/hwtracing/coresight/coresight-funnel.c | 9 +++------ drivers/hwtracing/coresight/coresight-replicator-qcom.c | 4 +--- drivers/hwtracing/coresight/coresight-replicator.c | 7 ++----- drivers/hwtracing/coresight/coresight-tmc.c | 9 +++------ drivers/hwtracing/coresight/coresight-tpiu.c | 9 +++------ drivers/hwtracing/coresight/coresight.c | 3 --- drivers/hwtracing/coresight/of_coresight.c | 1 - include/linux/amba/bus.h | 9 +++++++++ 11 files changed, 33 insertions(+), 45 deletions(-) (limited to 'include/linux') diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c index a2eb6bdeaafa..acbce79934d6 100644 --- a/drivers/hwtracing/coresight/coresight-etb10.c +++ b/drivers/hwtracing/coresight/coresight-etb10.c @@ -1,4 +1,6 @@ /* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Embedded Trace Buffer driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -12,7 +14,6 @@ #include #include -#include #include #include #include @@ -781,8 +782,4 @@ static struct amba_driver etb_driver = { .probe = etb_probe, .id_table = etb_ids, }; - -module_amba_driver(etb_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Embedded Trace Buffer driver"); +builtin_amba_driver(etb_driver); diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c index 0ba1a3981373..d83ab82672e4 100644 --- a/drivers/hwtracing/coresight/coresight-etm3x.c +++ b/drivers/hwtracing/coresight/coresight-etm3x.c @@ -1,4 +1,6 @@ /* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Program Flow Trace driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -11,7 +13,7 @@ */ #include -#include +#include #include #include #include @@ -38,6 +40,10 @@ #include "coresight-etm.h" #include "coresight-etm-perf.h" +/* + * Not really modular but using module_param is the easiest way to + * remain consistent with existing use cases for now. + */ static int boot_enable; module_param_named(boot_enable, boot_enable, int, S_IRUGO); @@ -912,8 +918,4 @@ static struct amba_driver etm_driver = { .probe = etm_probe, .id_table = etm_ids, }; - -module_amba_driver(etm_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Program Flow Trace driver"); +builtin_amba_driver(etm_driver); diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c index d0169ba7fbf2..1c59bd36834c 100644 --- a/drivers/hwtracing/coresight/coresight-etm4x.c +++ b/drivers/hwtracing/coresight/coresight-etm4x.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include @@ -2710,5 +2709,4 @@ static struct amba_driver etm4x_driver = { .probe = etm4_probe, .id_table = etm4_ids, }; - -module_amba_driver(etm4x_driver); +builtin_amba_driver(etm4x_driver); diff --git a/drivers/hwtracing/coresight/coresight-funnel.c b/drivers/hwtracing/coresight/coresight-funnel.c index fb2c679fbc44..0600ca30649d 100644 --- a/drivers/hwtracing/coresight/coresight-funnel.c +++ b/drivers/hwtracing/coresight/coresight-funnel.c @@ -1,4 +1,6 @@ /* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Funnel driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -11,7 +13,6 @@ */ #include -#include #include #include #include @@ -268,8 +269,4 @@ static struct amba_driver funnel_driver = { .probe = funnel_probe, .id_table = funnel_ids, }; - -module_amba_driver(funnel_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Funnel driver"); +builtin_amba_driver(funnel_driver); diff --git a/drivers/hwtracing/coresight/coresight-replicator-qcom.c b/drivers/hwtracing/coresight/coresight-replicator-qcom.c index 286f90b50989..700f710e4bfa 100644 --- a/drivers/hwtracing/coresight/coresight-replicator-qcom.c +++ b/drivers/hwtracing/coresight/coresight-replicator-qcom.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include @@ -198,5 +197,4 @@ static struct amba_driver replicator_driver = { .probe = replicator_probe, .id_table = replicator_ids, }; - -module_amba_driver(replicator_driver); +builtin_amba_driver(replicator_driver); diff --git a/drivers/hwtracing/coresight/coresight-replicator.c b/drivers/hwtracing/coresight/coresight-replicator.c index 0ce98903992c..4299c0569340 100644 --- a/drivers/hwtracing/coresight/coresight-replicator.c +++ b/drivers/hwtracing/coresight/coresight-replicator.c @@ -1,4 +1,6 @@ /* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Replicator driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -11,7 +13,6 @@ */ #include -#include #include #include #include @@ -166,8 +167,4 @@ static struct platform_driver replicator_driver = { .suppress_bind_attrs = true, }, }; - builtin_platform_driver(replicator_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Replicator driver"); diff --git a/drivers/hwtracing/coresight/coresight-tmc.c b/drivers/hwtracing/coresight/coresight-tmc.c index ac91b0b40ec2..1be191f5d39c 100644 --- a/drivers/hwtracing/coresight/coresight-tmc.c +++ b/drivers/hwtracing/coresight/coresight-tmc.c @@ -1,4 +1,6 @@ /* Copyright (c) 2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Trace Memory Controller driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -11,7 +13,6 @@ */ #include -#include #include #include #include @@ -782,8 +783,4 @@ static struct amba_driver tmc_driver = { .probe = tmc_probe, .id_table = tmc_ids, }; - -module_amba_driver(tmc_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Trace Memory Controller driver"); +builtin_amba_driver(tmc_driver); diff --git a/drivers/hwtracing/coresight/coresight-tpiu.c b/drivers/hwtracing/coresight/coresight-tpiu.c index 71582f6dfd4c..8fb09d9237ab 100644 --- a/drivers/hwtracing/coresight/coresight-tpiu.c +++ b/drivers/hwtracing/coresight/coresight-tpiu.c @@ -1,4 +1,6 @@ /* Copyright (c) 2011-2012, The Linux Foundation. All rights reserved. + * + * Description: CoreSight Trace Port Interface Unit driver * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -11,7 +13,6 @@ */ #include -#include #include #include #include @@ -218,8 +219,4 @@ static struct amba_driver tpiu_driver = { .probe = tpiu_probe, .id_table = tpiu_ids, }; - -module_amba_driver(tpiu_driver); - -MODULE_LICENSE("GPL v2"); -MODULE_DESCRIPTION("CoreSight Trace Port Interface Unit driver"); +builtin_amba_driver(tpiu_driver); diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index 6ec2b66af9ee..2ea5961092c1 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -11,7 +11,6 @@ */ #include -#include #include #include #include @@ -894,5 +893,3 @@ void coresight_unregister(struct coresight_device *csdev) device_unregister(&csdev->dev); } EXPORT_SYMBOL_GPL(coresight_unregister); - -MODULE_LICENSE("GPL v2"); diff --git a/drivers/hwtracing/coresight/of_coresight.c b/drivers/hwtracing/coresight/of_coresight.c index 3cc57c1e3b5d..b68da1888fd5 100644 --- a/drivers/hwtracing/coresight/of_coresight.c +++ b/drivers/hwtracing/coresight/of_coresight.c @@ -10,7 +10,6 @@ * GNU General Public License for more details. */ -#include #include #include #include diff --git a/include/linux/amba/bus.h b/include/linux/amba/bus.h index 9006c4e75cf7..3d8dcdd1aeae 100644 --- a/include/linux/amba/bus.h +++ b/include/linux/amba/bus.h @@ -163,4 +163,13 @@ struct amba_device name##_device = { \ #define module_amba_driver(__amba_drv) \ module_driver(__amba_drv, amba_driver_register, amba_driver_unregister) +/* + * builtin_amba_driver() - Helper macro for drivers that don't do anything + * special in driver initcall. This eliminates a lot of boilerplate. Each + * driver may only use this macro once, and calling it replaces the instance + * device_initcall(). + */ +#define builtin_amba_driver(__amba_drv) \ + builtin_driver(__amba_drv, amba_driver_register) + #endif -- cgit