* [PATCH] stackdepot: Use a raw spinlock in stack depot
@ 2021-05-04 2:43 Andi Kleen
2021-05-04 7:23 ` Dmitry Vyukov
0 siblings, 1 reply; 5+ messages in thread
From: Andi Kleen @ 2021-05-04 2:43 UTC (permalink / raw)
To: linux-kernel
Cc: Andi Kleen, Andrey Ryabinin, Alexander Potapenko,
Andrey Konovalov, Dmitry Vyukov, peterz, akpm
In some cases kasan with lockdep complains about its own use of the stack
depot lock. I think it happens when the kasan usage is nested inside a
raw spinlock and it happens to create a new stack depot.
Make the stackdepot lock raw too
Example trace:
[ 1.156154] ACPI Error: No handler or method for GPE 01, disabling event (20210105/evgpe-839)
[ 1.156235] =============================
[ 1.156238] [ BUG: Invalid wait context ]
[ 1.156242] 5.12.0-00071-gb34886074b65 #45 Not tainted
[ 1.156249] -----------------------------
[ 1.156252] swapper/0/1 is trying to lock:
[ 1.156258] ffffffff8535e158 (depot_lock){..-.}-{3:3}, at: stack_depot_save+0x162/0x450
[ 1.156288] other info that might help us debug this:
[ 1.156292] context-{5:5}
[ 1.156297] 3 locks held by swapper/0/1:
[ 1.156304] #0: ffff888100838658 (*(&acpi_gbl_gpe_lock)){....}-{3:3}, at: acpi_ev_detect_gpe+0xa8/0x3e6
[ 1.156333] #1: ffffffff85133dc0 (rcu_read_lock){....}-{1:3}, at: __queue_work+0xd5/0x1010
[ 1.156363] #2: ffff88815a04ec18 (&pool->lock){....}-{2:2}, at: __queue_work+0x258/0x1010
[ 1.156391] stack backtrace:
[ 1.156395] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.12.0-00071-gb34886074b65 #45
[ 1.156407] Call Trace:
[ 1.156413] dump_stack+0xfa/0x151
[ 1.156427] __lock_acquire.cold+0x366/0x41f
[ 1.156450] ? orc_find+0x233/0x3c0
[ 1.156462] ? __sanitizer_cov_trace_pc+0x1d/0x50
[ 1.156476] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
[ 1.156491] ? ret_from_fork+0x1f/0x30
[ 1.156503] ? deref_stack_reg+0x90/0x90
[ 1.156516] lock_acquire+0x194/0x690
[ 1.156528] ? stack_depot_save+0x162/0x450
[ 1.156541] ? rcu_read_unlock+0x40/0x40
[ 1.156553] ? arch_stack_walk+0x88/0xf0
[ 1.156566] ? ret_from_fork+0x1f/0x30
[ 1.156579] _raw_spin_lock_irqsave+0x3b/0x60
[ 1.156591] ? stack_depot_save+0x162/0x450
[ 1.156603] stack_depot_save+0x162/0x450
[ 1.156616] kasan_save_stack+0x32/0x40
[ 1.156629] ? kasan_save_stack+0x1b/0x40
[ 1.156642] ? kasan_record_aux_stack+0xbc/0xe0
[ 1.156653] ? insert_work+0x4b/0x2e0
[ 1.156665] ? __queue_work+0x4cf/0x1010
[ 1.156677] ? queue_work_on+0xb3/0xc0
[ 1.156689] ? acpi_os_execute+0x1ca/0x340
[ 1.156701] ? acpi_ev_gpe_dispatch+0x208/0x273
[ 1.156713] ? acpi_ev_detect_gpe+0x35e/0x3e6
[ 1.156724] ? acpi_ev_gpe_detect+0x269/0x334
[ 1.156736] ? acpi_update_all_gpes+0x1cf/0x206
[ 1.156748] ? acpi_scan_init+0x2a8/0x702
[ 1.156761] ? acpi_init+0x230/0x2ba
[ 1.156773] ? do_one_initcall+0xf0/0x540
[ 1.156784] ? kernel_init_freeable+0x38e/0x412
[ 1.156796] ? kernel_init+0x12/0x1cf
[ 1.156807] ? ret_from_fork+0x1f/0x30
[ 1.156818] ? kmem_cache_alloc_trace+0xf1/0x850
[ 1.156831] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
[ 1.156844] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
[ 1.156858] ? acpi_scan_init+0x2a8/0x702
[ 1.156870] ? do_one_initcall+0xf0/0x540
[ 1.156881] ? kernel_init+0x12/0x1cf
[ 1.156892] ? ret_from_fork+0x1f/0x30
[ 1.156903] ? _raw_spin_unlock+0x1f/0x30
[ 1.156915] ? lock_acquire+0x194/0x690
[ 1.156927] ? __queue_work+0x258/0x1010
[ 1.156940] ? rcu_read_unlock+0x40/0x40
[ 1.156952] ? lock_is_held_type+0x98/0x110
[ 1.156964] ? lock_is_held_type+0x98/0x110
[ 1.156977] ? rcu_read_lock_sched_held+0xa1/0xe0
[ 1.156991] kasan_record_aux_stack+0xbc/0xe0
[ 1.157003] insert_work+0x4b/0x2e0
[ 1.157016] __queue_work+0x4cf/0x1010
[ 1.157031] queue_work_on+0xb3/0xc0
[ 1.157044] acpi_os_execute+0x1ca/0x340
[ 1.157056] acpi_ev_gpe_dispatch+0x208/0x273
[ 1.157069] acpi_ev_detect_gpe+0x35e/0x3e6
[ 1.157082] ? acpi_ev_gpe_dispatch+0x273/0x273
[ 1.157096] ? lockdep_hardirqs_on_prepare+0x273/0x3e0
[ 1.157109] ? _raw_spin_unlock_irqrestore+0x2d/0x40
[ 1.157122] acpi_ev_gpe_detect+0x269/0x334
[ 1.157136] ? acpi_bus_init+0x7ee/0x7ee
[ 1.157148] acpi_update_all_gpes+0x1cf/0x206
[ 1.157161] ? acpi_get_gpe_device+0x182/0x182
[ 1.157174] ? acpi_get_table+0x11f/0x1f5
[ 1.157186] ? write_comp_data+0x2a/0x80
[ 1.157199] acpi_scan_init+0x2a8/0x702
[ 1.157212] ? acpi_match_madt+0xc4/0xc4
[ 1.157225] ? __sanitizer_cov_trace_pc+0x1d/0x50
[ 1.157239] ? __pci_mmcfg_init+0x91/0x21f
[ 1.157251] ? __sanitizer_cov_trace_pc+0x1d/0x50
[ 1.157265] acpi_init+0x230/0x2ba
[ 1.157277] ? acpi_bus_init+0x7ee/0x7ee
[ 1.157290] ? rcu_read_lock_bh_held+0xc0/0xc0
[ 1.157304] ? write_comp_data+0x2a/0x80
[ 1.157332] do_one_initcall+0xf0/0x540
[ 1.157357] ? perf_trace_initcall_level+0x3e0/0x3e0
[ 1.157370] ? rcu_read_lock_sched_held+0xa1/0xe0
[ 1.157384] ? rcu_read_lock_bh_held+0xc0/0xc0
[ 1.157398] ? __kmalloc+0x1ae/0x380
[ 1.157410] ? write_comp_data+0x2a/0x80
[ 1.157424] kernel_init_freeable+0x38e/0x412
[ 1.157437] ? rest_init+0x381/0x381
[ 1.157462] kernel_init+0x12/0x1cf
[ 1.157474] ret_from_fork+0x1f/0x30
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: peterz@infradead.org
Cc: akpm@linux-foundation.org
Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
lib/stackdepot.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 49f67a0c6e5d..df9179f4f441 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -71,7 +71,7 @@ static void *stack_slabs[STACK_ALLOC_MAX_SLABS];
static int depot_index;
static int next_slab_inited;
static size_t depot_offset;
-static DEFINE_SPINLOCK(depot_lock);
+static DEFINE_RAW_SPINLOCK(depot_lock);
static bool init_stack_slab(void **prealloc)
{
@@ -305,7 +305,7 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries,
prealloc = page_address(page);
}
- spin_lock_irqsave(&depot_lock, flags);
+ raw_spin_lock_irqsave(&depot_lock, flags);
found = find_stack(*bucket, entries, nr_entries, hash);
if (!found) {
@@ -329,7 +329,7 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries,
WARN_ON(!init_stack_slab(&prealloc));
}
- spin_unlock_irqrestore(&depot_lock, flags);
+ raw_spin_unlock_irqrestore(&depot_lock, flags);
exit:
if (prealloc) {
/* Nobody used this memory, ok to free it. */
--
2.25.4
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH] stackdepot: Use a raw spinlock in stack depot
2021-05-04 2:43 [PATCH] stackdepot: Use a raw spinlock in stack depot Andi Kleen
@ 2021-05-04 7:23 ` Dmitry Vyukov
2021-05-04 9:49 ` Peter Zijlstra
2021-05-04 15:34 ` Andi Kleen
0 siblings, 2 replies; 5+ messages in thread
From: Dmitry Vyukov @ 2021-05-04 7:23 UTC (permalink / raw)
To: Andi Kleen
Cc: LKML, Andrey Ryabinin, Alexander Potapenko, Andrey Konovalov,
Peter Zijlstra, Andrew Morton, kasan-dev
On Tue, May 4, 2021 at 4:44 AM Andi Kleen <ak@linux.intel.com> wrote:
>
> In some cases kasan with lockdep complains about its own use of the stack
> depot lock. I think it happens when the kasan usage is nested inside a
> raw spinlock and it happens to create a new stack depot.
>
> Make the stackdepot lock raw too
+kasan-dev
Hi Andi,
So why is this a false positive that we just need to silence?
I see LOCKDEP is saying we are doing something wrong, and your
description just describes how we are doing something wrong :)
If this is a special false positive case, it would be good to have a
comment on DEFINE_RAW_SPINLOCK explaining why we are using it.
I wonder why we never saw this on syzbot. Is it an RT kernel or some
other special config?
A similar issue was discussed recently for RT kernel:
https://groups.google.com/g/kasan-dev/c/MyHh8ov-ciU/m/nahiuqFLAQAJ
And I think it may be fixable in the same way -- make stackdepot not
allocate in contexts where it's not OK to allocate.
> Example trace:
>
> [ 1.156154] ACPI Error: No handler or method for GPE 01, disabling event (20210105/evgpe-839)
> [ 1.156235] =============================
> [ 1.156238] [ BUG: Invalid wait context ]
> [ 1.156242] 5.12.0-00071-gb34886074b65 #45 Not tainted
> [ 1.156249] -----------------------------
> [ 1.156252] swapper/0/1 is trying to lock:
> [ 1.156258] ffffffff8535e158 (depot_lock){..-.}-{3:3}, at: stack_depot_save+0x162/0x450
> [ 1.156288] other info that might help us debug this:
> [ 1.156292] context-{5:5}
> [ 1.156297] 3 locks held by swapper/0/1:
> [ 1.156304] #0: ffff888100838658 (*(&acpi_gbl_gpe_lock)){....}-{3:3}, at: acpi_ev_detect_gpe+0xa8/0x3e6
> [ 1.156333] #1: ffffffff85133dc0 (rcu_read_lock){....}-{1:3}, at: __queue_work+0xd5/0x1010
> [ 1.156363] #2: ffff88815a04ec18 (&pool->lock){....}-{2:2}, at: __queue_work+0x258/0x1010
> [ 1.156391] stack backtrace:
> [ 1.156395] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.12.0-00071-gb34886074b65 #45
> [ 1.156407] Call Trace:
> [ 1.156413] dump_stack+0xfa/0x151
> [ 1.156427] __lock_acquire.cold+0x366/0x41f
> [ 1.156450] ? orc_find+0x233/0x3c0
> [ 1.156462] ? __sanitizer_cov_trace_pc+0x1d/0x50
> [ 1.156476] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
> [ 1.156491] ? ret_from_fork+0x1f/0x30
> [ 1.156503] ? deref_stack_reg+0x90/0x90
> [ 1.156516] lock_acquire+0x194/0x690
> [ 1.156528] ? stack_depot_save+0x162/0x450
> [ 1.156541] ? rcu_read_unlock+0x40/0x40
> [ 1.156553] ? arch_stack_walk+0x88/0xf0
> [ 1.156566] ? ret_from_fork+0x1f/0x30
> [ 1.156579] _raw_spin_lock_irqsave+0x3b/0x60
> [ 1.156591] ? stack_depot_save+0x162/0x450
> [ 1.156603] stack_depot_save+0x162/0x450
> [ 1.156616] kasan_save_stack+0x32/0x40
> [ 1.156629] ? kasan_save_stack+0x1b/0x40
> [ 1.156642] ? kasan_record_aux_stack+0xbc/0xe0
> [ 1.156653] ? insert_work+0x4b/0x2e0
> [ 1.156665] ? __queue_work+0x4cf/0x1010
> [ 1.156677] ? queue_work_on+0xb3/0xc0
> [ 1.156689] ? acpi_os_execute+0x1ca/0x340
> [ 1.156701] ? acpi_ev_gpe_dispatch+0x208/0x273
> [ 1.156713] ? acpi_ev_detect_gpe+0x35e/0x3e6
> [ 1.156724] ? acpi_ev_gpe_detect+0x269/0x334
> [ 1.156736] ? acpi_update_all_gpes+0x1cf/0x206
> [ 1.156748] ? acpi_scan_init+0x2a8/0x702
> [ 1.156761] ? acpi_init+0x230/0x2ba
> [ 1.156773] ? do_one_initcall+0xf0/0x540
> [ 1.156784] ? kernel_init_freeable+0x38e/0x412
> [ 1.156796] ? kernel_init+0x12/0x1cf
> [ 1.156807] ? ret_from_fork+0x1f/0x30
> [ 1.156818] ? kmem_cache_alloc_trace+0xf1/0x850
> [ 1.156831] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
> [ 1.156844] ? lockdep_hardirqs_on_prepare+0x3e0/0x3e0
> [ 1.156858] ? acpi_scan_init+0x2a8/0x702
> [ 1.156870] ? do_one_initcall+0xf0/0x540
> [ 1.156881] ? kernel_init+0x12/0x1cf
> [ 1.156892] ? ret_from_fork+0x1f/0x30
> [ 1.156903] ? _raw_spin_unlock+0x1f/0x30
> [ 1.156915] ? lock_acquire+0x194/0x690
> [ 1.156927] ? __queue_work+0x258/0x1010
> [ 1.156940] ? rcu_read_unlock+0x40/0x40
> [ 1.156952] ? lock_is_held_type+0x98/0x110
> [ 1.156964] ? lock_is_held_type+0x98/0x110
> [ 1.156977] ? rcu_read_lock_sched_held+0xa1/0xe0
> [ 1.156991] kasan_record_aux_stack+0xbc/0xe0
> [ 1.157003] insert_work+0x4b/0x2e0
> [ 1.157016] __queue_work+0x4cf/0x1010
> [ 1.157031] queue_work_on+0xb3/0xc0
> [ 1.157044] acpi_os_execute+0x1ca/0x340
> [ 1.157056] acpi_ev_gpe_dispatch+0x208/0x273
> [ 1.157069] acpi_ev_detect_gpe+0x35e/0x3e6
> [ 1.157082] ? acpi_ev_gpe_dispatch+0x273/0x273
> [ 1.157096] ? lockdep_hardirqs_on_prepare+0x273/0x3e0
> [ 1.157109] ? _raw_spin_unlock_irqrestore+0x2d/0x40
> [ 1.157122] acpi_ev_gpe_detect+0x269/0x334
> [ 1.157136] ? acpi_bus_init+0x7ee/0x7ee
> [ 1.157148] acpi_update_all_gpes+0x1cf/0x206
> [ 1.157161] ? acpi_get_gpe_device+0x182/0x182
> [ 1.157174] ? acpi_get_table+0x11f/0x1f5
> [ 1.157186] ? write_comp_data+0x2a/0x80
> [ 1.157199] acpi_scan_init+0x2a8/0x702
> [ 1.157212] ? acpi_match_madt+0xc4/0xc4
> [ 1.157225] ? __sanitizer_cov_trace_pc+0x1d/0x50
> [ 1.157239] ? __pci_mmcfg_init+0x91/0x21f
> [ 1.157251] ? __sanitizer_cov_trace_pc+0x1d/0x50
> [ 1.157265] acpi_init+0x230/0x2ba
> [ 1.157277] ? acpi_bus_init+0x7ee/0x7ee
> [ 1.157290] ? rcu_read_lock_bh_held+0xc0/0xc0
> [ 1.157304] ? write_comp_data+0x2a/0x80
> [ 1.157332] do_one_initcall+0xf0/0x540
> [ 1.157357] ? perf_trace_initcall_level+0x3e0/0x3e0
> [ 1.157370] ? rcu_read_lock_sched_held+0xa1/0xe0
> [ 1.157384] ? rcu_read_lock_bh_held+0xc0/0xc0
> [ 1.157398] ? __kmalloc+0x1ae/0x380
> [ 1.157410] ? write_comp_data+0x2a/0x80
> [ 1.157424] kernel_init_freeable+0x38e/0x412
> [ 1.157437] ? rest_init+0x381/0x381
> [ 1.157462] kernel_init+0x12/0x1cf
> [ 1.157474] ret_from_fork+0x1f/0x30
>
> Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
> Cc: Alexander Potapenko <glider@google.com>
> Cc: Andrey Konovalov <andreyknvl@gmail.com>
> Cc: Dmitry Vyukov <dvyukov@google.com>
> Cc: peterz@infradead.org
> Cc: akpm@linux-foundation.org
> Signed-off-by: Andi Kleen <ak@linux.intel.com>
> ---
> lib/stackdepot.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> index 49f67a0c6e5d..df9179f4f441 100644
> --- a/lib/stackdepot.c
> +++ b/lib/stackdepot.c
> @@ -71,7 +71,7 @@ static void *stack_slabs[STACK_ALLOC_MAX_SLABS];
> static int depot_index;
> static int next_slab_inited;
> static size_t depot_offset;
> -static DEFINE_SPINLOCK(depot_lock);
> +static DEFINE_RAW_SPINLOCK(depot_lock);
>
> static bool init_stack_slab(void **prealloc)
> {
> @@ -305,7 +305,7 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries,
> prealloc = page_address(page);
> }
>
> - spin_lock_irqsave(&depot_lock, flags);
> + raw_spin_lock_irqsave(&depot_lock, flags);
>
> found = find_stack(*bucket, entries, nr_entries, hash);
> if (!found) {
> @@ -329,7 +329,7 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries,
> WARN_ON(!init_stack_slab(&prealloc));
> }
>
> - spin_unlock_irqrestore(&depot_lock, flags);
> + raw_spin_unlock_irqrestore(&depot_lock, flags);
> exit:
> if (prealloc) {
> /* Nobody used this memory, ok to free it. */
> --
> 2.25.4
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] stackdepot: Use a raw spinlock in stack depot
2021-05-04 7:23 ` Dmitry Vyukov
@ 2021-05-04 9:49 ` Peter Zijlstra
2021-05-04 15:34 ` Andi Kleen
1 sibling, 0 replies; 5+ messages in thread
From: Peter Zijlstra @ 2021-05-04 9:49 UTC (permalink / raw)
To: Dmitry Vyukov
Cc: Andi Kleen, LKML, Andrey Ryabinin, Alexander Potapenko,
Andrey Konovalov, Andrew Morton, kasan-dev
On Tue, May 04, 2021 at 09:23:34AM +0200, Dmitry Vyukov wrote:
> So why is this a false positive that we just need to silence?
No, it's a correctness issue for PREEMPT_RT.
> I see LOCKDEP is saying we are doing something wrong, and your
> description just describes how we are doing something wrong :)
> If this is a special false positive case, it would be good to have a
> comment on DEFINE_RAW_SPINLOCK explaining why we are using it.
Documentation/locking/locktypes.rst has the low-down IIRC
> I wonder why we never saw this on syzbot. Is it an RT kernel or some
> other special config?
IIRC the kernel isn't really PROVE_RAW_LOCK_NESTING=y clean yet, so
mostly these checks aren't on by default. printk() used to be a common
offender, but I've not checked the very latest printk status to see if
that got fixed meanwhile.
> A similar issue was discussed recently for RT kernel:
> https://groups.google.com/g/kasan-dev/c/MyHh8ov-ciU/m/nahiuqFLAQAJ
> And I think it may be fixable in the same way -- make stackdepot not
> allocate in contexts where it's not OK to allocate.
That would be preferable I think.
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] stackdepot: Use a raw spinlock in stack depot
2021-05-04 7:23 ` Dmitry Vyukov
2021-05-04 9:49 ` Peter Zijlstra
@ 2021-05-04 15:34 ` Andi Kleen
2021-05-04 17:17 ` Dmitry Vyukov
1 sibling, 1 reply; 5+ messages in thread
From: Andi Kleen @ 2021-05-04 15:34 UTC (permalink / raw)
To: Dmitry Vyukov
Cc: LKML, Andrey Ryabinin, Alexander Potapenko, Andrey Konovalov,
Peter Zijlstra, Andrew Morton, kasan-dev
> So why is this a false positive that we just need to silence?
> I see LOCKDEP is saying we are doing something wrong, and your
> description just describes how we are doing something wrong :)
> If this is a special false positive case, it would be good to have a
> comment on DEFINE_RAW_SPINLOCK explaining why we are using it.
>
> I wonder why we never saw this on syzbot. Is it an RT kernel or some
> other special config?
This happened in a special configuration that triggered ACPI errors at
boot time.
It's probably not something that is normally executed, as well as syzbot is
probably not exercising bootup anyways.
> A similar issue was discussed recently for RT kernel:
> https://groups.google.com/g/kasan-dev/c/MyHh8ov-ciU/m/nahiuqFLAQAJ
> And I think it may be fixable in the same way -- make stackdepot not
> allocate in contexts where it's not OK to allocate.
Yes that's a good idea. I've seen also other errors about the allocator
triggered
by stack depot being in the wrong context. Probably doing that would be
the right
fix. But I actually tried to switch depot to GFP_ATOMIC allocations
(from GFP_NOWAIT),
but it didn't help, so I'm not fully sure what needs to be changed.
-Andi
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] stackdepot: Use a raw spinlock in stack depot
2021-05-04 15:34 ` Andi Kleen
@ 2021-05-04 17:17 ` Dmitry Vyukov
0 siblings, 0 replies; 5+ messages in thread
From: Dmitry Vyukov @ 2021-05-04 17:17 UTC (permalink / raw)
To: Andi Kleen
Cc: LKML, Andrey Ryabinin, Alexander Potapenko, Andrey Konovalov,
Peter Zijlstra, Andrew Morton, kasan-dev
On Tue, May 4, 2021 at 5:34 PM Andi Kleen <ak@linux.intel.com> wrote:
> > So why is this a false positive that we just need to silence?
> > I see LOCKDEP is saying we are doing something wrong, and your
> > description just describes how we are doing something wrong :)
> > If this is a special false positive case, it would be good to have a
> > comment on DEFINE_RAW_SPINLOCK explaining why we are using it.
> >
> > I wonder why we never saw this on syzbot. Is it an RT kernel or some
> > other special config?
>
> This happened in a special configuration that triggered ACPI errors at
> boot time.
>
> It's probably not something that is normally executed, as well as syzbot is
>
> probably not exercising bootup anyways.
>
> > A similar issue was discussed recently for RT kernel:
> > https://groups.google.com/g/kasan-dev/c/MyHh8ov-ciU/m/nahiuqFLAQAJ
> > And I think it may be fixable in the same way -- make stackdepot not
> > allocate in contexts where it's not OK to allocate.
>
>
> Yes that's a good idea. I've seen also other errors about the allocator
> triggered
>
> by stack depot being in the wrong context. Probably doing that would be
> the right
>
> fix. But I actually tried to switch depot to GFP_ATOMIC allocations
> (from GFP_NOWAIT),
>
> but it didn't help, so I'm not fully sure what needs to be changed.
We may not allocate at all, see may_prealloc idea here:
https://groups.google.com/g/kasan-dev/c/MyHh8ov-ciU/m/k1LXBmonAQAJ
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-05-04 17:17 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-04 2:43 [PATCH] stackdepot: Use a raw spinlock in stack depot Andi Kleen
2021-05-04 7:23 ` Dmitry Vyukov
2021-05-04 9:49 ` Peter Zijlstra
2021-05-04 15:34 ` Andi Kleen
2021-05-04 17:17 ` Dmitry Vyukov
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.