Re: [PATCH v1 1/2] perf/x86: Avoid inadvertent casts to x86_hybrid_pmu

From: Mi, Dapeng

Date: Thu Mar 12 2026 - 05:46:01 EST



On 3/12/2026 4:31 PM, Peter Zijlstra wrote:
> On Wed, Mar 11, 2026 at 10:48:09PM -0700, Ian Rogers wrote:
>> The patch:
>> https://lore.kernel.org/lkml/20260311075201.2951073-2-dapeng1.mi@xxxxxxxxxxxxxxx/
>> showed it was pretty easy to accidentally cast non-x86 PMUs to
>> x86_hybrid_pmus. Add a BUG_ON for that case. Restructure is_x86_event
>> and add an is_x86_pmu to facilitate this.
>>
>> @@ -779,6 +795,7 @@ struct x86_hybrid_pmu {
>>
>> static __always_inline struct x86_hybrid_pmu *hybrid_pmu(struct pmu *pmu)
>> {
>> + BUG_ON(!is_x86_pmu(pmu));
>> return container_of(pmu, struct x86_hybrid_pmu, pmu);
>> }
> Given that hybrid_pmu will have PERF_PMU_CAP_EXTENDED_HW_TYPE, and we
> should really only use hyrid_pmu() on one of those, would not the
> simpler patch be so?
>
>
> diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
> index fad87d3c8b2c..13ec623617a9 100644
> --- a/arch/x86/events/perf_event.h
> +++ b/arch/x86/events/perf_event.h
> @@ -779,6 +779,7 @@ struct x86_hybrid_pmu {
>
> static __always_inline struct x86_hybrid_pmu *hybrid_pmu(struct pmu *pmu)
> {
> + BUG_ON(!(pmu->capabilities & PERF_PMU_CAP_EXTENDED_HW_TYPE));

It looks we can't add either !is_x86_pmu(pmu) or !(pmu->capabilities &
PERF_PMU_CAP_EXTENDED_HW_TYPE) here. hybrid_pmu() is called by the hybrid()
marco or other variants, and hybrid() macro is called in many places of the
intel_pmu_init(), like the update_pmu_cap() , but the flag
PERF_PMU_CAP_EXTENDED_HW_TYPE is still not set for the hybrid
pmu->capabilities until intel_pmu_init() ends and the hybrid pmus are
registered. Then it would cause the unexpected kernel crash.

[    1.945128] kernel BUG at arch/x86/events/intel/../perf_event.h:798!
[    1.946131] Oops: invalid opcode: 0000 [#1] SMP NOPTI
[    1.947127] CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Not tainted
7.0.0-rc3-perf-urgent-gc8b4b538960c #460 PREEMPT(full)
[    1.947127] Hardware name: Intel Corporation Panther Lake Client
Platform/PTL-UH LP5 T3 RVP1, BIOS PTLPFWI1.R00.3171.D00.2504220409 04/22/2025
[    1.947127] RIP: 0010:intel_pmu_init+0x25c9/0x5fd0
[    1.947127] Code: db 44 ff 4c 89 35 c7 da 44 ff 48 89 2d 80 da 44 ff e9
49 df ff ff 83 7a 68 04 0f 84 1b f9 ff ff f6 42 6d 01 0f 85 11 f9 ff ff
<0f> 0b 31 d2 48 89 df
[    1.947127] RSP: 0000:ffffd5dc800f7db8 EFLAGS: 00010246
[    1.947127] RAX: 0000000000000001 RBX: 00000000000abfff RCX:
0000000000000000
[    1.947127] RDX: ffff8f40856bc000 RSI: 0000000000000001 RDI:
00000000000000ff
[    1.947127] RBP: 0000000000000001 R08: ffffffffffffffff R09:
0000000000000004
[    1.947127] R10: ffffffffbd4e2500 R11: 0000000000000006 R12:
ffffffffbc26438b
[    1.947127] R13: 0000000000000000 R14: 0000000000000000 R15:
0000000000000000
[    1.947127] FS:  0000000000000000(0000) GS:ffff8f482214f000(0000)
knlGS:0000000000000000
[    1.947127] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    1.947127] CR2: ffff8f47ff7ff000 CR3: 00000004c1434001 CR4:
0000000000f70ef0
[    1.947127] PKRU: 55555554
[    1.947127] Call Trace:
[    1.947127]  <TASK>
[    1.947127]  ? __pfx_init_hw_perf_events+0x10/0x10
[    1.947127]  init_hw_perf_events+0x2af/0x4b0
[    1.947127]  ? __pfx_init_hw_perf_events+0x10/0x10
[    1.947127]  do_one_initcall+0x52/0x250
[    1.947127]  ? _raw_spin_unlock+0x18/0x40
[    1.947127]  ? __register_sysctl_table+0x143/0x1a0
[    1.947127]  kernel_init_freeable+0x21d/0x340
[    1.947127]  ? __pfx_kernel_init+0x10/0x10
[    1.947127]  kernel_init+0x1a/0x1c0
[    1.947127]  ret_from_fork+0xcb/0x1c0
[    1.947127]  ? __pfx_kernel_init+0x10/0x10
[    1.947127]  ret_from_fork_asm+0x1a/0x30
[    1.947127]  </TASK>
[    1.947127] Modules linked in:
[    1.947127] ---[ end trace 0000000000000000 ]---
[    1.948128] RIP: 0010:intel_pmu_init+0x25c9/0x5fd0
[    1.949128] Code: db 44 ff 4c 89 35 c7 da 44 ff 48 89 2d 80 da 44 ff e9
49 df ff ff 83 7a 68 04 0f 84 1b f9 ff ff f6 42 6d 01 0f 85 11 f9 ff ff
<0f> 0b 31 d2 48 89 df
[    1.950129] RSP: 0000:ffffd5dc800f7db8 EFLAGS: 00010246
[    1.951128] RAX: 0000000000000001 RBX: 00000000000abfff RCX:
0000000000000000
[    1.952128] RDX: ffff8f40856bc000 RSI: 0000000000000001 RDI:
00000000000000ff
[    1.953128] RBP: 0000000000000001 R08: ffffffffffffffff R09:
0000000000000004
[    1.954129] R10: ffffffffbd4e2500 R11: 0000000000000006 R12:
ffffffffbc26438b
[    1.955128] R13: 0000000000000000 R14: 0000000000000000 R15:
0000000000000000
[    1.956128] FS:  0000000000000000(0000) GS:ffff8f482214f000(0000)
knlGS:0000000000000000
[    1.957128] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    1.958128] CR2: ffff8f47ff7ff000 CR3: 00000004c1434001 CR4:
0000000000f70ef0
[    1.959128] PKRU: 55555554
[    1.960128] Kernel panic - not syncing: Attempted to kill init!
exitcode=0x0000000b

I'm not sure if we can move the flag PERF_PMU_CAP_EXTENDED_HW_TYPE setting
earlier and eventually find a good place to set the flag. Even it's
possible, but could be risky ... 

Ian, if you don't object, I would suggest to drop the bug_on(). I would
adopt other changes and add the is_x86_pmu() check in the
x86_pmu_has_rdpmc_user_disable() to fix the issue.

Thanks.

> return container_of(pmu, struct x86_hybrid_pmu, pmu);
> }
>