* v5.19-rc2-rt3: mm/kfence might_sleep() splat
@ 2022-06-18 9:34 Mike Galbraith
2022-06-24 9:05 ` Sebastian Andrzej Siewior
0 siblings, 1 reply; 4+ messages in thread
From: Mike Galbraith @ 2022-06-18 9:34 UTC (permalink / raw)
To: RT; +Cc: Sebastian Andrzej Siewior
I moved the prandom_u32_max() call in kfence_guarded_alloc() out from
under raw spinlock to shut this one up.
[ 1.128544] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:46
[ 1.128546] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 90, name: kworker/u16:3
[ 1.128547] preempt_count: 1, expected: 0
[ 1.128548] RCU nest depth: 1, expected: 1
[ 1.128549] CPU: 3 PID: 90 Comm: kworker/u16:3 Tainted: G W 5.19.0.g0639b59-master-rt #2 55e5fbd63d8381661776ddec390c2b764f305c0b
[ 1.128551] Hardware name: MEDION MS-7848/MS-7848, BIOS M7848W08.20C 09/23/2013
[ 1.128552] Workqueue: events_unbound async_run_entry_fn
[ 1.128556] Call Trace:
[ 1.128557] <TASK>
[ 1.128558] dump_stack_lvl+0x44/0x58
[ 1.128562] __might_resched+0x141/0x160
[ 1.128566] rt_spin_lock+0x2d/0x70
[ 1.128569] get_random_u32+0x45/0x100
[ 1.128572] ? stack_trace_save+0x4b/0x70
[ 1.128575] __kfence_alloc+0x3f4/0x6c0
[ 1.128579] ? __kfence_alloc+0xa8/0x6c0
[ 1.128581] ? kmem_cache_alloc_lru+0x1d8/0x220
[ 1.128582] ? xas_alloc+0x9b/0xc0
[ 1.128585] ? xas_create+0x20c/0x390
[ 1.128586] ? xas_store+0x52/0x5a0
[ 1.128587] ? __filemap_add_folio+0x189/0x5a0
[ 1.128590] ? filemap_add_folio+0x38/0xa0
[ 1.128591] ? __filemap_get_folio+0x1b0/0x580
[ 1.128592] ? pagecache_get_page+0x13/0x80
[ 1.128595] ? simple_write_begin+0x20/0x2d0
[ 1.128597] ? generic_perform_write+0xae/0x1e0
[ 1.128599] ? __generic_file_write_iter+0x141/0x180
[ 1.128600] ? generic_file_write_iter+0x5d/0xb0
[ 1.128601] ? __kernel_write+0x139/0x2f0
[ 1.128604] ? kernel_write+0x56/0x1a0
[ 1.128606] ? xwrite.constprop.8+0x35/0x8e
[ 1.128609] ? do_copy+0xee/0x13a
[ 1.128610] ? write_buffer+0x27/0x37
[ 1.128612] ? flush_buffer+0x34/0x8b
[ 1.128614] ? unxz+0x1b8/0x301
[ 1.128616] ? unpack_to_rootfs+0x17f/0x2ae
[ 1.128617] ? do_populate_rootfs+0x59/0x108
[ 1.128619] ? async_run_entry_fn+0x2b/0x110
[ 1.128620] ? process_one_work+0x21f/0x4a0
[ 1.128622] ? worker_thread+0x39/0x3d0
[ 1.128624] ? kthread+0x13e/0x160
[ 1.128625] ? ret_from_fork+0x1f/0x30
[ 1.128628] ? get_random_u32+0xbb/0x100
[ 1.128629] ? new_slab+0x288/0x4e0
[ 1.128631] ? rt_spin_unlock+0x13/0x40
[ 1.128633] ? ___slab_alloc+0x31a/0x660
[ 1.128635] ? xas_alloc+0x9b/0xc0
[ 1.128637] ? rt_spin_unlock+0x13/0x40
[ 1.128638] ? ___slab_alloc+0x31a/0x660
[ 1.128640] ? __d_alloc+0x29/0x1f0
[ 1.128642] ? ___slab_alloc+0x31a/0x660
[ 1.128644] ? slab_pre_alloc_hook.constprop.85+0x37/0x1c0
[ 1.128646] ? xas_alloc+0x9b/0xc0
[ 1.128647] kmem_cache_alloc_lru+0x1d8/0x220
[ 1.128649] xas_alloc+0x9b/0xc0
[ 1.128651] xas_create+0x20c/0x390
[ 1.128653] xas_store+0x52/0x5a0
[ 1.128654] ? xas_load+0x37/0x40
[ 1.128654] ? xa_get_order+0x9a/0xf0
[ 1.128655] __filemap_add_folio+0x189/0x5a0
[ 1.128657] ? scan_shadow_nodes+0x30/0x30
[ 1.128660] filemap_add_folio+0x38/0xa0
[ 1.128661] __filemap_get_folio+0x1b0/0x580
[ 1.128662] ? mntput_no_expire+0x4a/0x230
[ 1.128665] pagecache_get_page+0x13/0x80
[ 1.128667] simple_write_begin+0x20/0x2d0
[ 1.128669] generic_perform_write+0xae/0x1e0
[ 1.128671] __generic_file_write_iter+0x141/0x180
[ 1.128672] generic_file_write_iter+0x5d/0xb0
[ 1.128674] __kernel_write+0x139/0x2f0
[ 1.128676] kernel_write+0x56/0x1a0
[ 1.128678] xwrite.constprop.8+0x35/0x8e
[ 1.128680] ? initrd_load+0x3f/0x3f
[ 1.128682] do_copy+0xee/0x13a
[ 1.128684] ? write_buffer+0x27/0x37
[ 1.128685] write_buffer+0x27/0x37
[ 1.128687] flush_buffer+0x34/0x8b
[ 1.128689] ? write_buffer+0x37/0x37
[ 1.128690] unxz+0x1b8/0x301
[ 1.128692] ? initrd_load+0x3f/0x3f
[ 1.128693] ? unlzo+0x3cc/0x3cc
[ 1.128695] unpack_to_rootfs+0x17f/0x2ae
[ 1.128696] ? initrd_load+0x3f/0x3f
[ 1.128698] do_populate_rootfs+0x59/0x108
[ 1.128700] async_run_entry_fn+0x2b/0x110
[ 1.128701] process_one_work+0x21f/0x4a0
[ 1.128703] worker_thread+0x39/0x3d0
[ 1.128704] ? process_one_work+0x4a0/0x4a0
[ 1.128706] kthread+0x13e/0x160
[ 1.128707] ? kthread_complete_and_exit+0x20/0x20
[ 1.128709] ret_from_fork+0x1f/0x30
[ 1.128711] </TASK>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: v5.19-rc2-rt3: mm/kfence might_sleep() splat
2022-06-18 9:34 v5.19-rc2-rt3: mm/kfence might_sleep() splat Mike Galbraith
@ 2022-06-24 9:05 ` Sebastian Andrzej Siewior
2022-06-24 9:10 ` Marco Elver
0 siblings, 1 reply; 4+ messages in thread
From: Sebastian Andrzej Siewior @ 2022-06-24 9:05 UTC (permalink / raw)
To: Mike Galbraith
Cc: RT, Alexander Potapenko, Marco Elver, Dmitry Vyukov, kasan-dev
On 2022-06-18 11:34:51 [+0200], Mike Galbraith wrote:
> I moved the prandom_u32_max() call in kfence_guarded_alloc() out from
> under raw spinlock to shut this one up.
Care to send a patch? I don't even why kfence_metadata::lock is a
raw_spinlock_t. This is the case since the beginning of the code.
> [ 1.128544] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:46
> [ 1.128546] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 90, name: kworker/u16:3
> [ 1.128547] preempt_count: 1, expected: 0
> [ 1.128548] RCU nest depth: 1, expected: 1
> [ 1.128549] CPU: 3 PID: 90 Comm: kworker/u16:3 Tainted: G W 5.19.0.g0639b59-master-rt #2 55e5fbd63d8381661776ddec390c2b764f305c0b
> [ 1.128551] Hardware name: MEDION MS-7848/MS-7848, BIOS M7848W08.20C 09/23/2013
> [ 1.128552] Workqueue: events_unbound async_run_entry_fn
> [ 1.128556] Call Trace:
> [ 1.128557] <TASK>
> [ 1.128558] dump_stack_lvl+0x44/0x58
> [ 1.128562] __might_resched+0x141/0x160
> [ 1.128566] rt_spin_lock+0x2d/0x70
> [ 1.128569] get_random_u32+0x45/0x100
> [ 1.128575] __kfence_alloc+0x3f4/0x6c0
> [ 1.128647] kmem_cache_alloc_lru+0x1d8/0x220
> [ 1.128649] xas_alloc+0x9b/0xc0
> [ 1.128651] xas_create+0x20c/0x390
> [ 1.128653] xas_store+0x52/0x5a0
> [ 1.128655] __filemap_add_folio+0x189/0x5a0
> [ 1.128660] filemap_add_folio+0x38/0xa0
> [ 1.128661] __filemap_get_folio+0x1b0/0x580
> [ 1.128665] pagecache_get_page+0x13/0x80
> [ 1.128667] simple_write_begin+0x20/0x2d0
> [ 1.128669] generic_perform_write+0xae/0x1e0
> [ 1.128671] __generic_file_write_iter+0x141/0x180
> [ 1.128672] generic_file_write_iter+0x5d/0xb0
> [ 1.128674] __kernel_write+0x139/0x2f0
> [ 1.128676] kernel_write+0x56/0x1a0
> [ 1.128678] xwrite.constprop.8+0x35/0x8e
> [ 1.128682] do_copy+0xee/0x13a
> [ 1.128685] write_buffer+0x27/0x37
> [ 1.128687] flush_buffer+0x34/0x8b
> [ 1.128690] unxz+0x1b8/0x301
> [ 1.128695] unpack_to_rootfs+0x17f/0x2ae
> [ 1.128698] do_populate_rootfs+0x59/0x108
> [ 1.128700] async_run_entry_fn+0x2b/0x110
> [ 1.128701] process_one_work+0x21f/0x4a0
> [ 1.128703] worker_thread+0x39/0x3d0
> [ 1.128706] kthread+0x13e/0x160
> [ 1.128709] ret_from_fork+0x1f/0x30
> [ 1.128711] </TASK>
Sebastian
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: v5.19-rc2-rt3: mm/kfence might_sleep() splat
2022-06-24 9:05 ` Sebastian Andrzej Siewior
@ 2022-06-24 9:10 ` Marco Elver
2022-06-24 9:19 ` Sebastian Andrzej Siewior
0 siblings, 1 reply; 4+ messages in thread
From: Marco Elver @ 2022-06-24 9:10 UTC (permalink / raw)
To: Sebastian Andrzej Siewior
Cc: Mike Galbraith, RT, Alexander Potapenko, Dmitry Vyukov, kasan-dev
On Fri, 24 Jun 2022 at 11:05, Sebastian Andrzej Siewior
<bigeasy@linutronix.de> wrote:
>
> On 2022-06-18 11:34:51 [+0200], Mike Galbraith wrote:
> > I moved the prandom_u32_max() call in kfence_guarded_alloc() out from
> > under raw spinlock to shut this one up.
>
> Care to send a patch? I don't even why kfence_metadata::lock is a
> raw_spinlock_t. This is the case since the beginning of the code.
Because kfence_handle_page_fault() may be called from anywhere, incl.
other raw_spinlock critical sections. We have this problem with all
debugging tools where the bug may manifest anywhere.
A patch for it already exists in -mm:
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git/commit/?h=mm-hotfixes-stable&id=327b18b7aaed5de3b548212e3ab75133bf323759
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: v5.19-rc2-rt3: mm/kfence might_sleep() splat
2022-06-24 9:10 ` Marco Elver
@ 2022-06-24 9:19 ` Sebastian Andrzej Siewior
0 siblings, 0 replies; 4+ messages in thread
From: Sebastian Andrzej Siewior @ 2022-06-24 9:19 UTC (permalink / raw)
To: Marco Elver
Cc: Mike Galbraith, RT, Alexander Potapenko, Dmitry Vyukov, kasan-dev
On 2022-06-24 11:10:55 [+0200], Marco Elver wrote:
> On Fri, 24 Jun 2022 at 11:05, Sebastian Andrzej Siewior
> <bigeasy@linutronix.de> wrote:
> >
> > On 2022-06-18 11:34:51 [+0200], Mike Galbraith wrote:
> > > I moved the prandom_u32_max() call in kfence_guarded_alloc() out from
> > > under raw spinlock to shut this one up.
> >
> > Care to send a patch? I don't even why kfence_metadata::lock is a
> > raw_spinlock_t. This is the case since the beginning of the code.
>
> Because kfence_handle_page_fault() may be called from anywhere, incl.
> other raw_spinlock critical sections. We have this problem with all
> debugging tools where the bug may manifest anywhere.
Oh thank you. I had some vague memory of this but could find anything.
> A patch for it already exists in -mm:
> https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git/commit/?h=mm-hotfixes-stable&id=327b18b7aaed5de3b548212e3ab75133bf323759
Thanks for the pointer.
Sebastian
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2022-06-24 9:19 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-06-18 9:34 v5.19-rc2-rt3: mm/kfence might_sleep() splat Mike Galbraith
2022-06-24 9:05 ` Sebastian Andrzej Siewior
2022-06-24 9:10 ` Marco Elver
2022-06-24 9:19 ` Sebastian Andrzej Siewior
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.