| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
As discussed with Dae R. Jeong and Hillf Danton here [1] the sendmsg()
function in isotp.c might get into a race condition when restoring the
former tx.state from the old_state.
Remove the old_state concept and implement proper locking for the
ISOTP_IDLE transitions in isotp_sendmsg(), inspired by a
simplification idea from Hillf Danton.
Introduce a new tx.state ISOTP_SHUTDOWN and use the same locking
mechanism from isotp_release() which resolves a potential race between
isotp_sendsmg() and isotp_release().
[1] https://lore.kernel.org/linux-can/ZB%2F93xJxq%2FBUqAgG@dragonet
v1: https://lore.kernel.org/all/[email protected]
v2: https://lore.kernel.org/all/[email protected]
take care of signal interrupts for wait_event_interruptible() in
isotp_release()
v3: https://lore.kernel.org/all/[email protected]
take care of signal interrupts for wait_event_interruptible() in
isotp_sendmsg() in the wait_tx_done case
v4: https://lore.kernel.org/all/[email protected]
take care of signal interrupts for wait_event_interruptible() in
isotp_sendmsg() in ALL cases
Cc: Dae R. Jeong <[email protected]>
Cc: Hillf Danton <[email protected]>
Signed-off-by: Oliver Hartkopp <[email protected]>
Fixes: 4f027cba8216 ("can: isotp: split tx timer into transmission and timeout")
Link: https://lore.kernel.org/all/[email protected]
Cc: [email protected]
[mkl: rephrase commit message]
Signed-off-by: Marc Kleine-Budde <[email protected]>
|
|
git://anongit.freedesktop.org/drm/drm-intel into drm-fixes
drm/i915 fixes for v6.3-rc6:
- Fix DP MST DSC M/N calculation to use compressed bpp
- Fix racy use-after-free in perf ioctl
- Fix context runtime accounting
- Fix handling of GT reset during HuC loading
- Fix use of unsigned vm_fault_t for error values
Signed-off-by: Daniel Vetter <[email protected]>
From: Jani Nikula <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
When using select()/poll()/epoll() with a non-blocking ISOTP socket to
wait for when non-blocking write is possible, a false EPOLLOUT event
is sometimes returned. This can happen at least after sending a
message which must be split to multiple CAN frames.
The reason is that isotp_sendmsg() returns -EAGAIN when tx.state is
not equal to ISOTP_IDLE and this behavior is not reflected in
datagram_poll(), which is used in isotp_ops.
This is fixed by introducing ISOTP-specific poll function, which
suppresses the EPOLLOUT events in that case.
v2: https://lore.kernel.org/all/[email protected]
v1: https://lore.kernel.org/all/[email protected]
https://lore.kernel.org/all/[email protected]
Signed-off-by: Michal Sojka <[email protected]>
Reported-by: Jakub Jira <[email protected]>
Tested-by: Oliver Hartkopp <[email protected]>
Acked-by: Oliver Hartkopp <[email protected]>
Fixes: e057dd3fc20f ("can: add ISO 15765-2:2016 transport protocol")
Link: https://lore.kernel.org/all/[email protected]
Cc: [email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>
|
|
isotp.c was still using sock_recv_timestamp() which does not provide
control messages to detect dropped PDUs in the receive path.
Fixes: e057dd3fc20f ("can: add ISO 15765-2:2016 transport protocol")
Signed-off-by: Oliver Hartkopp <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Cc: [email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>
|
|
In the j1939_tp_tx_dat_new() function, an out-of-bounds memory access
could occur during the memcpy() operation if the size of skb->cb is
larger than the size of struct j1939_sk_buff_cb. This is because the
memcpy() operation uses the size of skb->cb, leading to a read beyond
the struct j1939_sk_buff_cb.
Updated the memcpy() operation to use the size of struct
j1939_sk_buff_cb instead of the size of skb->cb. This ensures that the
memcpy() operation only reads the memory within the bounds of struct
j1939_sk_buff_cb, preventing out-of-bounds memory access.
Additionally, add a BUILD_BUG_ON() to check that the size of skb->cb
is greater than or equal to the size of struct j1939_sk_buff_cb. This
ensures that the skb->cb buffer is large enough to hold the
j1939_sk_buff_cb structure.
Fixes: 9d71dd0c7009 ("can: add support of SAE J1939 protocol")
Reported-by: Shuangpeng Bai <[email protected]>
Tested-by: Shuangpeng Bai <[email protected]>
Signed-off-by: Oleksij Rempel <[email protected]>
Link: https://groups.google.com/g/syzkaller/c/G_LL-C3plRs/m/-8xCi6dCAgAJ
Link: https://lore.kernel.org/all/[email protected]
Cc: [email protected]
[mkl: rephrase commit message]
Signed-off-by: Marc Kleine-Budde <[email protected]>
|
|
The same task check in perf_event_set_output has some potential issues
for some usages.
For the current perf code, there is a problem if using of
perf_event_open() to have multiple samples getting into the same mmap’d
memory when they are both attached to the same process.
https://lore.kernel.org/all/[email protected]/
Because the event->ctx is not ready when the perf_event_set_output() is
invoked in the perf_event_open().
Besides the above issue, before the commit bd2756811766 ("perf: Rewrite
core context handling"), perf record can errors out when sampling with
a hardware event and a software event as below.
$ perf record -e cycles,dummy --per-thread ls
failed to mmap with 22 (Invalid argument)
That's because that prior to the commit a hardware event and a software
event are from different task context.
The problem should be a long time issue since commit c3f00c70276d
("perk: Separate find_get_context() from event initialization").
The task struct is stored in the event->hw.target for each per-thread
event. It is a more reliable way to determine whether two events are
attached to the same task.
The event->hw.target was also introduced several years ago by the
commit 50f16a8bf9d7 ("perf: Remove type specific target pointers"). It
can not only be used to fix the issue with the current code, but also
back port to fix the issues with an older kernel.
Note: The event->hw.target was introduced later than commit
c3f00c70276d. The patch may cannot be applied between the commit
c3f00c70276d and commit 50f16a8bf9d7. Anybody that wants to back-port
this at that period may have to find other solutions.
Fixes: c3f00c70276d ("perf: Separate find_get_context() from event initialization")
Signed-off-by: Kan Liang <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Reviewed-by: Zhengjun Xing <[email protected]>
Link: https://lkml.kernel.org/r/[email protected]
|
|
Thomas reported that offlining CPUs spends a lot of time in
synchronize_rcu() as called from perf_pmu_migrate_context() even though
he's not actually using uncore events.
Turns out, the thing is unconditionally waiting for RCU, even if there's
no actual events to migrate.
Fixes: 0cda4c023132 ("perf: Introduce perf_pmu_migrate_context()")
Reported-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Tested-by: Thomas Gleixner <[email protected]>
Reviewed-by: Thomas Gleixner <[email protected]>
Reviewed-by: Paul E. McKenney <[email protected]>
Link: https://lkml.kernel.org/r/[email protected]
|
|
Wait for VPU to be idle in ivpu_pm_suspend_cb() before powering off
the device, so jobs are not lost and TDRs are not triggered after
resume.
Fixes: 852be13f3bd3 ("accel/ivpu: Add PM support")
Signed-off-by: Stanislaw Gruszka <[email protected]>
Reviewed-by: Jeffrey Hugo <[email protected]>
Signed-off-by: Jacek Lawrynowicz <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
Currently job->done_fence is added to every BO handle within a job. If job
handle (command buffer) is shared between multiple submits, KMD will add
the fence in each of them. Then bo_wait_ioctl() executed on command buffer
will exit only when all jobs containing that handle are done.
This creates deadlock scenario for user mode driver in case when job handle
is added as dependency of another job, because bo_wait_ioctl() of first job
will wait until second job finishes, and second job can not finish before
first one.
Having fences added only to job buffer handle allows user space to execute
bo_wait_ioctl() on the job even if it's handle is submitted with other job.
Fixes: cd7272215c44 ("accel/ivpu: Add command buffer submission logic")
Signed-off-by: Karol Wachowski <[email protected]>
Signed-off-by: Stanislaw Gruszka <[email protected]>
Reviewed-by: Jeffrey Hugo <[email protected]>
Signed-off-by: Jacek Lawrynowicz <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
The kernel command line ftrace_boot_snapshot by itself is supposed to
trigger a snapshot at the end of boot up of the main top level trace
buffer. A ftrace_boot_snapshot=foo will do the same for an instance called
foo that was created by trace_instance=foo,...
The logic was broken where if ftrace_boot_snapshot was by itself, it would
trigger a snapshot for all instances that had tracing enabled, regardless
if it asked for a snapshot or not.
When a snapshot is requested for a buffer, the buffer's
tr->allocated_snapshot is set to true. Use that to know if a trace buffer
wants a snapshot at boot up or not.
Since the top level buffer is part of the ftrace_trace_arrays list,
there's no reason to treat it differently than the other buffers. Just
iterate the list if ftrace_boot_snapshot was specified.
Link: https://lkml.kernel.org/r/[email protected]
Cc: [email protected]
Cc: Masami Hiramatsu <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Ross Zwisler <[email protected]>
Fixes: 9c1c251d670bc ("tracing: Allow boot instances to have snapshot buffers")
Signed-off-by: Steven Rostedt (Google) <[email protected]>
|
|
appropriate instance
If a trace instance has a failure with its snapshot code, the error
message is to be written to that instance's buffer. But currently, the
message is written to the top level buffer. Worse yet, it may also disable
the top level buffer and not the instance that had the issue.
Link: https://lkml.kernel.org/r/[email protected]
Cc: [email protected]
Cc: Masami Hiramatsu <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Ross Zwisler <[email protected]>
Fixes: 2824f50332486 ("tracing: Make the snapshot trigger work with instances")
Signed-off-by: Steven Rostedt (Google) <[email protected]>
|
|
Non-GSO TCP packets whose SKBs' linear portion did not include the
entire TCP header were not populating the first Tx descriptor with
as many bytes as the vNIC expected. This change ensures that all
TCP packets populate the first descriptor with the correct number of
bytes.
Fixes: 893ce44df565 ("gve: Add basic driver framework for Compute Engine Virtual NIC")
Signed-off-by: Shailend Chand <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
syzbot reported a data-race in data-race in netlink_recvmsg() [1]
Indeed, netlink_recvmsg() can be run concurrently,
and netlink_dump() also needs protection.
[1]
BUG: KCSAN: data-race in netlink_recvmsg / netlink_recvmsg
read to 0xffff888141840b38 of 8 bytes by task 23057 on cpu 0:
netlink_recvmsg+0xea/0x730 net/netlink/af_netlink.c:1988
sock_recvmsg_nosec net/socket.c:1017 [inline]
sock_recvmsg net/socket.c:1038 [inline]
__sys_recvfrom+0x1ee/0x2e0 net/socket.c:2194
__do_sys_recvfrom net/socket.c:2212 [inline]
__se_sys_recvfrom net/socket.c:2208 [inline]
__x64_sys_recvfrom+0x78/0x90 net/socket.c:2208
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
write to 0xffff888141840b38 of 8 bytes by task 23037 on cpu 1:
netlink_recvmsg+0x114/0x730 net/netlink/af_netlink.c:1989
sock_recvmsg_nosec net/socket.c:1017 [inline]
sock_recvmsg net/socket.c:1038 [inline]
____sys_recvmsg+0x156/0x310 net/socket.c:2720
___sys_recvmsg net/socket.c:2762 [inline]
do_recvmmsg+0x2e5/0x710 net/socket.c:2856
__sys_recvmmsg net/socket.c:2935 [inline]
__do_sys_recvmmsg net/socket.c:2958 [inline]
__se_sys_recvmmsg net/socket.c:2951 [inline]
__x64_sys_recvmmsg+0xe2/0x160 net/socket.c:2951
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
value changed: 0x0000000000000000 -> 0x0000000000001000
Reported by Kernel Concurrency Sanitizer on:
CPU: 1 PID: 23037 Comm: syz-executor.2 Not tainted 6.3.0-rc4-syzkaller-00195-g5a57b48fdfcb #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/02/2023
Fixes: 9063e21fb026 ("netlink: autosize skb lengthes")
Reported-by: syzbot <[email protected]>
Signed-off-by: Eric Dumazet <[email protected]>
Reviewed-by: Simon Horman <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
If the number of lanes was forced and then subsequently the user
omits this parameter, the ksettings->lanes is reset. The driver
should then reset the number of lanes to the device's default
for the specified speed.
However, although the ksettings->lanes is set to 0, the mod variable
is not set to true to indicate the driver and userspace should be
notified of the changes.
The consequence is that the same ethtool operation will produce
different results based on the initial state.
If the initial state is:
$ ethtool swp1 | grep -A 3 'Speed: '
Speed: 500000Mb/s
Lanes: 2
Duplex: Full
Auto-negotiation: on
then executing 'ethtool -s swp1 speed 50000 autoneg off' will yield:
$ ethtool swp1 | grep -A 3 'Speed: '
Speed: 500000Mb/s
Lanes: 2
Duplex: Full
Auto-negotiation: off
While if the initial state is:
$ ethtool swp1 | grep -A 3 'Speed: '
Speed: 500000Mb/s
Lanes: 1
Duplex: Full
Auto-negotiation: off
executing the same 'ethtool -s swp1 speed 50000 autoneg off' results in:
$ ethtool swp1 | grep -A 3 'Speed: '
Speed: 500000Mb/s
Lanes: 1
Duplex: Full
Auto-negotiation: off
This patch fixes this behavior. Omitting lanes will always results in
the driver choosing the default lane width for the chosen speed. In this
scenario, regardless of the initial state, the end state will be, e.g.,
$ ethtool swp1 | grep -A 3 'Speed: '
Speed: 500000Mb/s
Lanes: 2
Duplex: Full
Auto-negotiation: off
Fixes: 012ce4dd3102 ("ethtool: Extend link modes settings uAPI with lanes")
Signed-off-by: Andy Roulin <[email protected]>
Reviewed-by: Danielle Ratson <[email protected]>
Reviewed-by: Ido Schimmel <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Kuniyuki Iwashima says:
====================
raw/ping: Fix locking in /proc/net/{raw,icmp}.
The first patch fixes a NULL deref for /proc/net/raw and second one fixes
the same issue for ping sockets.
The first patch also converts hlist_nulls to hlist, but this is because
the current code uses sk_nulls_for_each() for lockless readers, instead
of sk_nulls_for_each_rcu() which adds memory barrier, but raw sockets
does not use the nulls marker nor SLAB_TYPESAFE_BY_RCU in the first place.
OTOH, the ping sockets already uses sk_nulls_for_each_rcu(), and such
conversion can be posted later for net-next.
====================
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
After commit dbca1596bbb0 ("ping: convert to RCU lookups, get rid
of rwlock"), we use RCU for ping sockets, but we should use spinlock
for /proc/net/icmp to avoid a potential NULL deref mentioned in
the previous patch.
Let's go back to using spinlock there.
Note we can convert ping sockets to use hlist instead of hlist_nulls
because we do not use SLAB_TYPESAFE_BY_RCU for ping sockets.
Fixes: dbca1596bbb0 ("ping: convert to RCU lookups, get rid of rwlock")
Signed-off-by: Kuniyuki Iwashima <[email protected]>
Reviewed-by: Eric Dumazet <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Dae R. Jeong reported a NULL deref in raw_get_next() [0].
It seems that the repro was running these sequences in parallel so
that one thread was iterating on a socket that was being freed in
another netns.
unshare(0x40060200)
r0 = syz_open_procfs(0x0, &(0x7f0000002080)='net/raw\x00')
socket$inet_icmp_raw(0x2, 0x3, 0x1)
pread64(r0, &(0x7f0000000000)=""/10, 0xa, 0x10000000007f)
After commit 0daf07e52709 ("raw: convert raw sockets to RCU"), we
use RCU and hlist_nulls_for_each_entry() to iterate over SOCK_RAW
sockets. However, we should use spinlock for slow paths to avoid
the NULL deref.
Also, SOCK_RAW does not use SLAB_TYPESAFE_BY_RCU, and the slab object
is not reused during iteration in the grace period. In fact, the
lockless readers do not check the nulls marker with get_nulls_value().
So, SOCK_RAW should use hlist instead of hlist_nulls.
Instead of adding an unnecessary barrier by sk_nulls_for_each_rcu(),
let's convert hlist_nulls to hlist and use sk_for_each_rcu() for
fast paths and sk_for_each() and spinlock for /proc/net/raw.
[0]:
general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
CPU: 2 PID: 20952 Comm: syz-executor.0 Not tainted 6.2.0-g048ec869bafd-dirty #7
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014
RIP: 0010:read_pnet include/net/net_namespace.h:383 [inline]
RIP: 0010:sock_net include/net/sock.h:649 [inline]
RIP: 0010:raw_get_next net/ipv4/raw.c:974 [inline]
RIP: 0010:raw_get_idx net/ipv4/raw.c:986 [inline]
RIP: 0010:raw_seq_start+0x431/0x800 net/ipv4/raw.c:995
Code: ef e8 33 3d 94 f7 49 8b 6d 00 4c 89 ef e8 b7 65 5f f7 49 89 ed 49 83 c5 98 0f 84 9a 00 00 00 48 83 c5 c8 48 89 e8 48 c1 e8 03 <42> 80 3c 30 00 74 08 48 89 ef e8 00 3d 94 f7 4c 8b 7d 00 48 89 ef
RSP: 0018:ffffc9001154f9b0 EFLAGS: 00010206
RAX: 0000000000000005 RBX: 1ffff1100302c8fd RCX: 0000000000000000
RDX: 0000000000000028 RSI: ffffc9001154f988 RDI: ffffc9000f77a338
RBP: 0000000000000029 R08: ffffffff8a50ffb4 R09: fffffbfff24b6bd9
R10: fffffbfff24b6bd9 R11: 0000000000000000 R12: ffff88801db73b78
R13: fffffffffffffff9 R14: dffffc0000000000 R15: 0000000000000030
FS: 00007f843ae8e700(0000) GS:ffff888063700000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000055bb9614b35f CR3: 000000003c672000 CR4: 00000000003506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
seq_read_iter+0x4c6/0x10f0 fs/seq_file.c:225
seq_read+0x224/0x320 fs/seq_file.c:162
pde_read fs/proc/inode.c:316 [inline]
proc_reg_read+0x23f/0x330 fs/proc/inode.c:328
vfs_read+0x31e/0xd30 fs/read_write.c:468
ksys_pread64 fs/read_write.c:665 [inline]
__do_sys_pread64 fs/read_write.c:675 [inline]
__se_sys_pread64 fs/read_write.c:672 [inline]
__x64_sys_pread64+0x1e9/0x280 fs/read_write.c:672
do_syscall_x64 arch/x86/entry/common.c:51 [inline]
do_syscall_64+0x4e/0xa0 arch/x86/entry/common.c:82
entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x478d29
Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f843ae8dbe8 EFLAGS: 00000246 ORIG_RAX: 0000000000000011
RAX: ffffffffffffffda RBX: 0000000000791408 RCX: 0000000000478d29
RDX: 000000000000000a RSI: 0000000020000000 RDI: 0000000000000003
RBP: 00000000f477909a R08: 0000000000000000 R09: 0000000000000000
R10: 000010000000007f R11: 0000000000000246 R12: 0000000000791740
R13: 0000000000791414 R14: 0000000000791408 R15: 00007ffc2eb48a50
</TASK>
Modules linked in:
---[ end trace 0000000000000000 ]---
RIP: 0010:read_pnet include/net/net_namespace.h:383 [inline]
RIP: 0010:sock_net include/net/sock.h:649 [inline]
RIP: 0010:raw_get_next net/ipv4/raw.c:974 [inline]
RIP: 0010:raw_get_idx net/ipv4/raw.c:986 [inline]
RIP: 0010:raw_seq_start+0x431/0x800 net/ipv4/raw.c:995
Code: ef e8 33 3d 94 f7 49 8b 6d 00 4c 89 ef e8 b7 65 5f f7 49 89 ed 49 83 c5 98 0f 84 9a 00 00 00 48 83 c5 c8 48 89 e8 48 c1 e8 03 <42> 80 3c 30 00 74 08 48 89 ef e8 00 3d 94 f7 4c 8b 7d 00 48 89 ef
RSP: 0018:ffffc9001154f9b0 EFLAGS: 00010206
RAX: 0000000000000005 RBX: 1ffff1100302c8fd RCX: 0000000000000000
RDX: 0000000000000028 RSI: ffffc9001154f988 RDI: ffffc9000f77a338
RBP: 0000000000000029 R08: ffffffff8a50ffb4 R09: fffffbfff24b6bd9
R10: fffffbfff24b6bd9 R11: 0000000000000000 R12: ffff88801db73b78
R13: fffffffffffffff9 R14: dffffc0000000000 R15: 0000000000000030
FS: 00007f843ae8e700(0000) GS:ffff888063700000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f92ff166000 CR3: 000000003c672000 CR4: 00000000003506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Fixes: 0daf07e52709 ("raw: convert raw sockets to RCU")
Reported-by: syzbot <[email protected]>
Reported-by: Dae R. Jeong <[email protected]>
Link: https://lore.kernel.org/netdev/ZCA2mGV_cmq7lIfV@dragonet/
Signed-off-by: Kuniyuki Iwashima <[email protected]>
Reviewed-by: Eric Dumazet <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Pick up the fixes (first 6 patches) from the DOE rework series from
Lukas for v6.3-rc.
Link: https://lore.kernel.org/all/[email protected]/
|
|
One motivation for mapping range registers to decoder objects is
to use those settings for region autodiscovery.
The need to map a region for devices programmed to use range registers
is especially urgent now that the kernel no longer routes "Soft
Reserved" ranges in the memory map to device-dax by default. The CXL
memory range loses all access mechanisms.
Complete the implementation by marking the DPA reservation and setting
the endpoint-decoder state to signal autodiscovery. Note that the
default settings of ways=1 and granularity=4096 set in cxl_decode_init()
do not need to be updated.
Fixes: 09d09e04d2fc ("cxl/dax: Create dax devices for CXL RAM regions")
Tested-by: Dave Jiang <[email protected]>
Tested-by: Gregory Price <[email protected]>
Link: https://lore.kernel.org/r/168012575521.221280.14177293493678527326.stgit@dwillia2-xfh.jf.intel.com
Reviewed-by: Dave Jiang <[email protected]>
Signed-off-by: Dan Williams <[email protected]>
|
|
Recall that range register emulation seeks to treat the 2 potential
range registers as Linux CXL "decoder" objects. The number of range
registers can be 1 or 2, while HDM decoder ranges can include more than
2.
Be careful not to confuse DVSEC range count with HDM capability decoder
count. Commit to range register earlier in devm_cxl_setup_hdm().
Otherwise, a device with more HDM decoders than range registers can set
@cxlhdm->decoder_count to an invalid value.
Avoid introducing a forward declaration by just moving the definition of
should_emulate_decoders() earlier in the file. should_emulate_decoders()
is unchanged.
Tested-by: Dave Jiang <[email protected]>
Fixes: d7a2153762c7 ("cxl/hdm: Add emulation when HDM decoders are not committed")
Reviewed-by: Jonathan Cameron <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/168012574932.221280.15944705098679646436.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
Each time the contents of a given HPA are potentially changed in a cache
incoherent manner the CXL core sets CXL_REGION_F_INCOHERENT to
invalidate CPU caches before the region is used.
Successful invocation of attach_target() indicates that DPA has been
newly assigned to a given HPA in the dynamic region creation flow.
However, attach_target() is also reused in the autodiscovery flow where
the region was activated by platform firmware. In that case there is no
need to invalidate caches because that region is already in active use
and nothing about the autodiscovery flow modifies the HPA-to-DPA
relationship.
In the autodiscovery case cxl_region_attach() exits early after
determining the endpoint decoder is already correctly attached to the
region.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Fan Ni <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/168002858817.50647.1217607907088920888.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
RCDs (CXL memory devices that link train without VH capability and show
up as root complex integrated endpoints), hide the presence of the link
between the endpoint and the host-bridge. The CXL region setup/teardown
paths assume that a link hop is present and go looking for at least one
'struct cxl_port' instance between the CXL root port-object and an
endpoint port-object leading to crashes of the form:
BUG: kernel NULL pointer dereference, address: 0000000000000008
[..]
RIP: 0010:cxl_region_setup_targets+0x3e9/0xae0 [cxl_core]
[..]
Call Trace:
<TASK>
cxl_region_attach+0x46c/0x7a0 [cxl_core]
cxl_create_region+0x20b/0x270 [cxl_core]
cxl_mock_mem_probe+0x641/0x800 [cxl_mock_mem]
platform_probe+0x5b/0xb0
Detect RCDs explicitly and skip walking the non-existent port hierarchy
between root and endpoint in that case.
While this has been a problem since:
commit 0a19bfc8de93 ("cxl/port: Add RCD endpoint port enumeration")
...it becomes a more reliable crash scenario with the new autodiscovery
implementation.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Ira Weiny <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/168002858268.50647.728091521032131326.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
The find_cxl_root() helper is used to lookup root decoders and other CXL
platform topology information for a given endpoint. It turns out that
for RCDs it has never worked. The result of find_cxl_root(&cxlmd->dev)
is always NULL for the RCH topology case because it expects to find a
cxl_port at the host-bridge. RCH topologies only have the root cxl_port
object with the host-bridge as a dport. While there are no reports of
this being a problem to date, by inspection region enumeration should
crash as a result of this problem, and it does in a local unit test for
this scenario.
However, an observation that ever since:
commit f17b558d6663 ("cxl/pmem: Refactor nvdimm device registration, delete the workqueue")
...all callers of find_cxl_root() occur after the memdev connection to
the port topology has been established. That means that find_cxl_root()
can be simplified to a walk of the endpoint port topology to the root.
Switch to that arrangement which also fixes the RCD bug.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Jonathan Cameron <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/168002857715.50647.344876437247313909.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
If the driver is allowed to enable memory operation itself then it can
also turn on HDM decoder support at will.
With this the second call to cxl_setup_hdm_decoder_from_dvsec(), when
an HDM decoder is not committed, is not needed.
Fixes: b777e9bec960 ("cxl/hdm: Emulate HDM decoder from DVSEC range registers")
Link: http://lore.kernel.org/r/[email protected]
Reported-by: Jonathan Cameron <[email protected]>
Tested-by: Jonathan Cameron <[email protected]>
Reviewed-by: Jonathan Cameron <[email protected]>
Reviewed-by: Fan Ni <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Link: https://lore.kernel.org/r/167703068474.185722.664126485486344246.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
devm_cxl_setup_emulated_hdm() reallocates an instance of @cxlhdm that
was already allocated at the start of devm_cxl_setup_hdm(). Only one is
needed and devm_cxl_setup_emulated_hdm() does not do enough to warrant
being an explicit helper.
Fixes: 4474ce565ee4 ("cxl/hdm: Create emulated cxl_hdm for devices that do not have HDM decoders")
Tested-by: Dave Jiang <[email protected]>
Reviewed-by: Dave Jiang <[email protected]>
Reviewed-by: Jonathan Cameron <[email protected]>
Link: https://lore.kernel.org/r/167703067936.185722.7908921750127154779.stgit@dwillia2-xfh.jf.intel.com
Link: https://lore.kernel.org/r/168012574357.221280.5001364964799725366.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <[email protected]>
|
|
Polling needs a bio with a valid bi_bdev, but neither of those are
guaranteed for polled driver requests. Make request based polling
directly use blk-mq's polling function instead.
When executing a request from a polled hctx, we know the request's
cookie, and that it's from a live blk-mq queue that supports polling, so
we can safely skip everything that bio_poll provides.
Cc: [email protected]
Reported-by: Martin Belanger <[email protected]>
Reported-by: Daniel Wagner <[email protected]>
Signed-off-by: Keith Busch <[email protected]>
Tested-by: Daniel Wagner <[email protected]>
Revieded-by: Daniel Wagner <[email protected]>
Reviewed-by: Chaitanya Kulkarni <[email protected]>
Reviewed-by: Sagi Grimberg <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Tested-by: Shin'ichiro Kawasaki <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
Pull kvm fixes from Paolo Bonzini:
"PPC:
- Hide KVM_CAP_IRQFD_RESAMPLE if XIVE is enabled
s390:
- Fix handling of external interrupts in protected guests
x86:
- Resample the pending state of IOAPIC interrupts when unmasking them
- Fix usage of Hyper-V "enlightened TLB" on AMD
- Small fixes to real mode exceptions
- Suppress pending MMIO write exits if emulator detects exception
Documentation:
- Fix rST syntax"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
docs: kvm: x86: Fix broken field list
KVM: PPC: Make KVM_CAP_IRQFD_RESAMPLE platform dependent
KVM: s390: pv: fix external interruption loop not always detected
KVM: nVMX: Do not report error code when synthesizing VM-Exit from Real Mode
KVM: x86: Clear "has_error_code", not "error_code", for RM exception injection
KVM: x86: Suppress pending MMIO write exits if emulator detects exception
KVM: x86/ioapic: Resample the pending state of an IRQ when unmasking
KVM: irqfd: Make resampler_list an RCU list
KVM: SVM: Flush Hyper-V TLB when required
|
|
Initialization must be completed before calling _vdpa_register_device()
since it can connect the device to the vDPA bus, so requests can arrive
after that call.
So for example vdpasim_net_work(), which uses the net->*_stats variables,
can be scheduled before they are initialized.
Let's move _vdpa_register_device() to the end of vdpasim_net_dev_add()
and add a comment to avoid future issues.
Fixes: 0899774cb360 ("vdpa_sim_net: vendor satistics")
Signed-off-by: Stefano Garzarella <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Jason Wang <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux
Pull nfsd fixes from Chuck Lever:
- Fix a crash and a resource leak in NFSv4 COMPOUND processing
- Fix issues with AUTH_SYS credential handling
- Try again to address an NFS/NFSD/SUNRPC build dependency regression
* tag 'nfsd-6.3-5' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
NFSD: callback request does not use correct credential for AUTH_SYS
NFS: Remove "select RPCSEC_GSS_KRB5
sunrpc: only free unix grouplist after RCU settles
nfsd: call op_release, even when op_func returns an error
NFSD: Avoid calling OPDESC() with ops->opnum == OP_ILLEGAL
|
|
Add a missing ":" to fix a broken field list.
Signed-off-by: Takahiro Itazuri <[email protected]>
Fixes: ba7bb663f554 ("KVM: x86: Provide per VM capability for disabling PMU virtualization")
Message-Id: <[email protected]>
Signed-off-by: Paolo Bonzini <[email protected]>
|
|
Code that passes a 32-bit constant into cmpxchg() produces a harmless
sparse warning because of the truncation in the branch that is not taken:
fs/erofs/zdata.c: note: in included file (through /home/arnd/arm-soc/arch/arm/include/asm/cmpxchg.h, /home/arnd/arm-soc/arch/arm/include/asm/atomic.h, /home/arnd/arm-soc/include/linux/atomic.h, ...):
include/asm-generic/cmpxchg-local.h:29:33: warning: cast truncates bits from constant value (5f0ecafe becomes fe)
include/asm-generic/cmpxchg-local.h:33:34: warning: cast truncates bits from constant value (5f0ecafe becomes cafe)
include/asm-generic/cmpxchg-local.h:29:33: warning: cast truncates bits from constant value (5f0ecafe becomes fe)
include/asm-generic/cmpxchg-local.h:30:42: warning: cast truncates bits from constant value (5f0edead becomes ad)
include/asm-generic/cmpxchg-local.h:33:34: warning: cast truncates bits from constant value (5f0ecafe becomes cafe)
include/asm-generic/cmpxchg-local.h:34:44: warning: cast truncates bits from constant value (5f0edead becomes dead)
This was reported as a regression to Matt's recent __generic_cmpxchg_local
patch, though this patch only added more warnings on top of the ones
that were already there.
Rewording the truncation to use an explicit bitmask instead of a cast
to a smaller type avoids the warning but otherwise leaves the code
unchanged.
I had another look at why the cast is even needed for atomic_cmpxchg(),
and as Matt describes the problem here is that atomic_t contains a
signed 'int', but cmpxchg() takes an 'unsigned long' argument, and
converting between the two leads to a 64-bit sign-extension of
negative 32-bit atomics.
I checked the other implementations of arch_cmpxchg() and did not find
any others that run into the same problem as __generic_cmpxchg_local(),
but it's easy to be on the safe side here and always convert the
signed int into an unsigned int when calling arch_cmpxchg(), as this
will work even when any of the arch_cmpxchg() implementations run
into the same problem.
Fixes: 624654152284 ("locking/atomic: cmpxchg: Make __generic_cmpxchg_local compare against zero-extended 'old' value")
Reviewed-by: Matt Evans <[email protected]>
Signed-off-by: Arnd Bergmann <[email protected]>
|
|
Copy the forced type casts from the normal MMIO accessors to suppress
the sparse warnings that point out __raw_readl() returns a native endian
word (just like readl()).
Signed-off-by: Vladimir Oltean <[email protected]>
Signed-off-by: Arnd Bergmann <[email protected]>
|
|
Commit c1d55d50139b ("asm-generic/io.h: Fix sparse warnings on
big-endian architectures") missed fixing the 64-bit accessors.
Arnd explains in the attached link why the casts are necessary, even if
__raw_readq() and __raw_writeq() do not take endian-specific types.
Link: https://lore.kernel.org/lkml/[email protected]/
Suggested-by: Arnd Bergmann <[email protected]>
Signed-off-by: Vladimir Oltean <[email protected]>
Reviewed-by: Jonathan Cameron <[email protected]>
Signed-off-by: Arnd Bergmann <[email protected]>
|
|
The existing pKVM code attempts to advertise CSV2/3 using values
initialized to 0, but never set. To advertise CSV2/3 to protected
guests, pass the CSV2/3 values to hyp when initializing hyp's
view of guests' ID_AA64PFR0_EL1.
Similar to non-protected KVM, these are system-wide, rather than
per cpu, for simplicity.
Fixes: 6c30bfb18d0b ("KVM: arm64: Add handlers for protected VM System Registers")
Signed-off-by: Fuad Tabba <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Oliver Upton <[email protected]>
|
|
Reset the FDIR counters when FDIR inits. Without this patch,
when VF initializes or resets, all the FDIR counters are not
cleaned, which may cause unexpected behaviors for future FDIR
rule create (e.g., rule conflict).
Fixes: 1f7ea1cd6a37 ("ice: Enable FDIR Configure for AVF")
Signed-off-by: Junfeng Guo <[email protected]>
Signed-off-by: Lingyu Liu <[email protected]>
Tested-by: Rafal Romanowski <[email protected]>
Signed-off-by: Tony Nguyen <[email protected]>
|
|
When adding a FDIR filter, if ice_vc_fdir_set_irq_ctx returns failure,
the inserted fdir entry will not be removed and if ice_vc_fdir_write_fltr
returns failure, the fdir context info for irq handler will not be cleared
which may lead to inconsistent or memory leak issue. This patch refines
failure cases to resolve this issue.
Fixes: 1f7ea1cd6a37 ("ice: Enable FDIR Configure for AVF")
Signed-off-by: Simei Su <[email protected]>
Tested-by: Rafal Romanowski <[email protected]>
Signed-off-by: Tony Nguyen <[email protected]>
|
|
The right place to add the debugfs create is in
setup_driver() and remove it in teardown_driver().
Current code adds the debugfs when creating the device but resetting a
device will remove the debugfs subtree and subsequent set_driver will
not be able to create the files since the debugfs pointer is NULL.
Fixes: 294221004322 ("vdpa/mlx5: Add debugfs subtree")
Signed-off-by: Eli Cohen <[email protected]>
v3 -> v4:
Fix error flow in setup_driver()
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Jason Wang <[email protected]>
|
|
We need to have a unique chardev for each data path, else the chardevs
will collide and qemu will die with this message:
qemu-system-x86_64: -device
virtserialport,bus=virtio-serial0.0,nr=2,chardev=charchannel0,
id=channel1,name=trace-path-cpu0:
Property 'virtserialport.chardev' can't take value 'charchannel0':
Device 'charchannel0' is in use
Signed-off-by: Ross Zwisler <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
|
|
We normally clear the endpoint then unmap LUNs so the devices are fully
shutdown when the LUN is unmapped, but it's legal to unmap before
clearing. If the user does that while TMFs are running then we can end
up crashing.
vhost_scsi_port_unlink assumes that the LUN's tmf struct will always be on
the tmf_queue list. However, if a TMF is running then it will have been
removed while it's executing. If we do a LUN unmap at this time, then
we assume the entry is on the list and just start accessing it and free
it.
This fixes the bug by just allocating the vhost_scsi_tmf struct when it's
needed like is done with the se_tmr struct that's needed when we submit
the TMF. In this path perf is not an issue and we can use GFP_KERNEL
since it won't swing directly back on us, so we don't need to preallocate
the struct.
Signed-off-by: Mike Christie <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
|
|
If vhost_scsi_setup_vq_cmds fails we leave the tpg->vhost_scsi pointer
set. If the device is freed and then the user unmaps the LUN, the call to
vhost_scsi_port_unlink -> vhost_scsi_hotunplug will see the that
tpg->vhost_scsi is still set and try to use it.
This has us clear the vhost_scsi pointer in the failure path. It also
has us take tv_tpg_mutex in this failure path, because tv_tpg_vhost_count
is accessed under this mutex in vhost_scsi_drop_nexus and in the future
we will want to serialize access to tpg->vhost_scsi with that mutex
instead of the vhost_scsi_mutex.
Signed-off-by: Mike Christie <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
|
|
When the kernel is built without support for zoned block devices,
virtio-blk probe needs to error out any host-managed device scans
to prevent such devices from appearing in the system as non-zoned.
The current virtio-blk code simply bypasses all ZBD checks if
CONFIG_BLK_DEV_ZONED is not defined and this leads to host-managed
block devices being presented as non-zoned in the OS. This is one of
the main problems this patch series is aimed to fix.
In this patch, make VIRTIO_BLK_F_ZONED feature defined even when
CONFIG_BLK_DEV_ZONED is not. This change makes the code compliant with
the voted revision of virtio-blk ZBD spec. Modify the probe code to
look at the situation when VIRTIO_BLK_F_ZONED is negotiated in a kernel
that is built without ZBD support. In this case, the code checks
the zoned model of the device and fails the probe is the device
is host-managed.
The patch also adds the comment to clarify that the call to perform
the zoned device probe is correctly placed after virtio_device ready().
Fixes: 95bfec41bd3d ("virtio-blk: add support for zoned block devices")
Cc: [email protected]
Signed-off-by: Dmitry Fomichev <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
|
|
The merged patch series to support zoned block devices in virtio-blk
is not the most up to date version. The merged patch can be found at
https://lore.kernel.org/linux-block/[email protected]/
but the latest and reviewed version is
https://lore.kernel.org/linux-block/[email protected]/
The reason is apparently that the correct mailing lists and
maintainers were not copied.
The differences between the two are mostly cleanups, but there is one
change that is very important in terms of compatibility with the
approved virtio-zbd specification.
Before it was approved, the OASIS virtio spec had a change in
VIRTIO_BLK_T_ZONE_APPEND request layout that is not reflected in the
current virtio-blk driver code. In the running code, the status is
the first byte of the in-header that is followed by some pad bytes
and the u64 that carries the sector at which the data has been written
to the zone back to the driver, aka the append sector.
This layout turned out to be problematic for implementing in QEMU and
the request status byte has been eventually made the last byte of the
in-header. The current code doesn't expect that and this causes the
append sector value always come as zero to the block layer. This needs
to be fixed ASAP.
Fixes: 95bfec41bd3d ("virtio-blk: add support for zoned block devices")
Cc: [email protected]
Signed-off-by: Dmitry Fomichev <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
|
|
Currently callback request does not use the credential specified in
CREATE_SESSION if the security flavor for the back channel is AUTH_SYS.
Problem was discovered by pynfs 4.1 DELEG5 and DELEG7 test with error:
DELEG5 st_delegation.testCBSecParms : FAILURE
expected callback with uid, gid == 17, 19, got 0, 0
Signed-off-by: Dai Ngo <[email protected]>
Reviewed-by: Jeff Layton <[email protected]>
Fixes: 8276c902bbe9 ("SUNRPC: remove uid and gid from struct auth_cred")
Signed-off-by: Chuck Lever <[email protected]>
|
|
If CONFIG_CRYPTO=n (e.g. arm/shmobile_defconfig):
WARNING: unmet direct dependencies detected for RPCSEC_GSS_KRB5
Depends on [n]: NETWORK_FILESYSTEMS [=y] && SUNRPC [=y] && CRYPTO [=n]
Selected by [y]:
- NFS_V4 [=y] && NETWORK_FILESYSTEMS [=y] && NFS_FS [=y]
As NFSv4 can work without crypto enabled, remove the RPCSEC_GSS_KRB5
dependency altogether.
Trond says:
> It is possible to use the NFSv4.1 client with just AUTH_SYS, and
> in fact there are plenty of people out there using only that. The
> fact that RFC5661 gets its knickers in a twist about RPCSEC_GSS
> support is largely irrelevant to those people.
>
> The other issue is that ’select’ enforces the strict dependency
> that if the NFS client is compiled into the kernel, then the
> RPCSEC_GSS and kerberos code needs to be compiled in as well: they
> cannot exist as modules.
Fixes: e57d06527738 ("NFS & NFSD: Update GSS dependencies")
Reported-by: kernel test robot <[email protected]>
Reported-by: Niklas Söderlund <[email protected]>
Suggested-by: Trond Myklebust <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
|
|
While the unix_gid object is rcu-freed, the group_info list that it
contains is not. Ensure that we only put the group list reference once
we are really freeing the unix_gid object.
Reported-by: Zhi Li <[email protected]>
Link: https://bugzilla.redhat.com/show_bug.cgi?id=2183056
Signed-off-by: Jeff Layton <[email protected]>
Fixes: fd5d2f78261b ("SUNRPC: Make server side AUTH_UNIX use lockless lookups")
Signed-off-by: Chuck Lever <[email protected]>
|
|
stmmac_reinit_queues() fails to fix up the RX hash. Even if the number
of channels gets restricted, the output of `ethtool -x' indicates that
all RX queues are used:
$ ethtool -l enp0s29f2
Channel parameters for enp0s29f2:
Pre-set maximums:
RX: 8
TX: 8
Other: n/a
Combined: n/a
Current hardware settings:
RX: 8
TX: 8
Other: n/a
Combined: n/a
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 8 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]
Fix this by setting the indirection table according to the number
of specified queues. The result is now as expected:
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 0 1 2 0 1
8: 2 0 1 2 0 1 2 0
[...]
Tested on Intel Elkhart Lake.
Fixes: 0366f7e06a6b ("net: stmmac: add ethtool support for get/set channels")
Signed-off-by: Corinna Vinschen <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>
|
|
If batch->end is 0 then setting npfns[0] before computing the new value of
pfns will fail to adjust the pfn and result in various page accounting
corruptions. It should be ordered after.
This seems to result in various kinds of page meta-data corruption related
failures:
WARNING: CPU: 1 PID: 527 at mm/gup.c:75 try_grab_folio+0x503/0x740
Modules linked in:
CPU: 1 PID: 527 Comm: repro Not tainted 6.3.0-rc2-eeac8ede1755+ #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
RIP: 0010:try_grab_folio+0x503/0x740
Code: e3 01 48 89 de e8 6d c1 dd ff 48 85 db 0f 84 7c fe ff ff e8 4f bf dd ff 49 8d 47 ff 48 89 45 d0 e9 73 fe ff ff e8 3d bf dd ff <0f> 0b 31 db e9 d0 fc ff ff e8 2f bf dd ff 48 8b 5d c8 31 ff 48 89
RSP: 0018:ffffc90000f37908 EFLAGS: 00010046
RAX: 0000000000000000 RBX: 00000000fffffc02 RCX: ffffffff81504c26
RDX: 0000000000000000 RSI: ffff88800d030000 RDI: 0000000000000002
RBP: ffffc90000f37948 R08: 000000000003ca24 R09: 0000000000000008
R10: 000000000003ca00 R11: 0000000000000023 R12: ffffea000035d540
R13: 0000000000000001 R14: 0000000000000000 R15: ffffea000035d540
FS: 00007fecbf659740(0000) GS:ffff88807dd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000200011c3 CR3: 000000000ef66006 CR4: 0000000000770ee0
PKRU: 55555554
Call Trace:
<TASK>
internal_get_user_pages_fast+0xd32/0x2200
pin_user_pages_fast+0x65/0x90
pfn_reader_user_pin+0x376/0x390
pfn_reader_next+0x14a/0x7b0
pfn_reader_first+0x140/0x1b0
iopt_area_fill_domain+0x74/0x210
iopt_table_add_domain+0x30e/0x6e0
iommufd_device_selftest_attach+0x7f/0x140
iommufd_test+0x10ff/0x16f0
iommufd_fops_ioctl+0x206/0x330
__x64_sys_ioctl+0x10e/0x160
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x72/0xdc
Cc: <[email protected]>
Fixes: f394576eb11d ("iommufd: PFN handling for iopt_pages")
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Kevin Tian <[email protected]>
Reported-by: Pengfei Xu <[email protected]>
Tested-by: Pengfei Xu <[email protected]>
Signed-off-by: Jason Gunthorpe <[email protected]>
|
|
syzkaller found that the calculation of batch_last_index should use
'start_index' since at input to this function the batch is either empty or
it has already been adjusted to cross any accesses so it will start at the
point we are unmapping from.
Getting this wrong causes the unmap to run over the end of the pages
which corrupts pages that were never mapped. In most cases this triggers
the num pinned debugging:
WARNING: CPU: 0 PID: 557 at drivers/iommu/iommufd/pages.c:294 __iopt_area_unfill_domain+0x152/0x560
Modules linked in:
CPU: 0 PID: 557 Comm: repro Not tainted 6.3.0-rc2-eeac8ede1755 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
RIP: 0010:__iopt_area_unfill_domain+0x152/0x560
Code: d2 0f ff 44 8b 64 24 54 48 8b 44 24 48 31 ff 44 89 e6 48 89 44 24 38 e8 fc d3 0f ff 45 85 e4 0f 85 eb 01 00 00 e8 0e d2 0f ff <0f> 0b e8 07 d2 0f ff 48 8b 44 24 38 89 5c 24 58 89 18 8b 44 24 54
RSP: 0018:ffffc9000108baf0 EFLAGS: 00010246
RAX: 0000000000000000 RBX: 00000000ffffffff RCX: ffffffff821e3f85
RDX: 0000000000000000 RSI: ffff88800faf0000 RDI: 0000000000000002
RBP: ffffc9000108bd18 R08: 000000000003ca25 R09: 0000000000000014
R10: 000000000003ca00 R11: 0000000000000024 R12: 0000000000000004
R13: 0000000000000801 R14: 00000000000007ff R15: 0000000000000800
FS: 00007f3499ce1740(0000) GS:ffff88807dc00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000243 CR3: 00000000179c2001 CR4: 0000000000770ef0
PKRU: 55555554
Call Trace:
<TASK>
iopt_area_unfill_domain+0x32/0x40
iopt_table_remove_domain+0x23f/0x4c0
iommufd_device_selftest_detach+0x3a/0x90
iommufd_selftest_destroy+0x55/0x70
iommufd_object_destroy_user+0xce/0x130
iommufd_destroy+0xa2/0xc0
iommufd_fops_ioctl+0x206/0x330
__x64_sys_ioctl+0x10e/0x160
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x72/0xdc
Also add some useful WARN_ON sanity checks.
Cc: <[email protected]>
Fixes: 8d160cd4d506 ("iommufd: Algorithms for PFN storage")
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Kevin Tian <[email protected]>
Reported-by: Pengfei Xu <[email protected]>
Tested-by: Pengfei Xu <[email protected]>
Signed-off-by: Jason Gunthorpe <[email protected]>
|
|
syzkaller found that setting up a map with a user VA that wraps past zero
can trigger WARN_ONs, particularly from pin_user_pages weirdly returning 0
due to invalid arguments.
Prevent creating a pages with a uptr and size that would math overflow.
WARNING: CPU: 0 PID: 518 at drivers/iommu/iommufd/pages.c:793 pfn_reader_user_pin+0x2e6/0x390
Modules linked in:
CPU: 0 PID: 518 Comm: repro Not tainted 6.3.0-rc2-eeac8ede1755+ #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
RIP: 0010:pfn_reader_user_pin+0x2e6/0x390
Code: b1 11 e9 25 fe ff ff e8 28 e4 0f ff 31 ff 48 89 de e8 2e e6 0f ff 48 85 db 74 0a e8 14 e4 0f ff e9 4d ff ff ff e8 0a e4 0f ff <0f> 0b bb f2 ff ff ff e9 3c ff ff ff e8 f9 e3 0f ff ba 01 00 00 00
RSP: 0018:ffffc90000f9fa30 EFLAGS: 00010246
RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffffff821e2b72
RDX: 0000000000000000 RSI: ffff888014184680 RDI: 0000000000000002
RBP: ffffc90000f9fa78 R08: 00000000000000ff R09: 0000000079de6f4e
R10: ffffc90000f9f790 R11: ffff888014185418 R12: ffffc90000f9fc60
R13: 0000000000000002 R14: ffff888007879800 R15: 0000000000000000
FS: 00007f4227555740(0000) GS:ffff88807dc00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000043 CR3: 000000000e748005 CR4: 0000000000770ef0
PKRU: 55555554
Call Trace:
<TASK>
pfn_reader_next+0x14a/0x7b0
? interval_tree_double_span_iter_update+0x11a/0x140
pfn_reader_first+0x140/0x1b0
iopt_pages_rw_slow+0x71/0x280
? __this_cpu_preempt_check+0x20/0x30
iopt_pages_rw_access+0x2b2/0x5b0
iommufd_access_rw+0x19f/0x2f0
iommufd_test+0xd11/0x16f0
? write_comp_data+0x2f/0x90
iommufd_fops_ioctl+0x206/0x330
__x64_sys_ioctl+0x10e/0x160
? __pfx_iommufd_fops_ioctl+0x10/0x10
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x72/0xdc
Cc: <[email protected]>
Fixes: 8d160cd4d506 ("iommufd: Algorithms for PFN storage")
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Kevin Tian <[email protected]>
Reported-by: Pengfei Xu <[email protected]>
Tested-by: Pengfei Xu <[email protected]>
Signed-off-by: Jason Gunthorpe <[email protected]>
|
|
In the am65_cpsw_nuss_probe() function's cleanup path, the call to
of_platform_device_destroy() for the common->mdio_dev device is invoked
unconditionally. It is possible that either the MDIO node is not present
in the device-tree, or the MDIO node is disabled in the device-tree. In
both these cases, the MDIO device is not created, resulting in a NULL
pointer dereference when the of_platform_device_destroy() function is
invoked on the common->mdio_dev device on the cleanup path.
Fix this by ensuring that the common->mdio_dev device exists, before
attempting to invoke of_platform_device_destroy().
Fixes: a45cfcc69a25 ("net: ethernet: ti: am65-cpsw-nuss: use of_platform_device_create() for mdio")
Signed-off-by: Siddharth Vadapalli <[email protected]>
Reviewed-by: Roger Quadros <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>
|