Re: [patch V2 32/52] x86/fpu/xstate: Sanitize handling of independent features

[Liang, Kan]

On 6/14/2021 11:44 AM, Thomas Gleixner wrote:
> The copy functions for the independent features are horribly named and the
> supervisor and independent part is just overengineered.
>
> The point is that the supplied mask has either to be a subset of the
> independent feature or a subset of the task->fpu.xstate managed features.
>
> Rewrite it so it checks check for invalid overlaps of these areas in the
> caller supplied feature mask. Rename it so it follows the new naming
> convention for these operations. Mop up the function documentation.
>
> This allows to use that function for other purposes as well.
>
> Suggested-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Cc: Kan Liang <kan.liang@xxxxxxxxxxxxxxx>
> ---
>   arch/x86/events/intel/lbr.c       |    6 +-
>   arch/x86/include/asm/fpu/xstate.h |    5 +-
>   arch/x86/kernel/fpu/xstate.c      |   93 +++++++++++++++++++-------------------
>   3 files changed, 53 insertions(+), 51 deletions(-)
>
> --- a/arch/x86/events/intel/lbr.c
> +++ b/arch/x86/events/intel/lbr.c
> @@ -491,7 +491,7 @@ static void intel_pmu_arch_lbr_xrstors(v
>   {
>   	struct x86_perf_task_context_arch_lbr_xsave *task_ctx = ctx;
>   
> -	copy_kernel_to_independent_supervisor(&task_ctx->xsave, XFEATURE_MASK_LBR);
> +	xrstors_from_kernel(&task_ctx->xsave, XFEATURE_MASK_LBR);
>   }
>   
>   static __always_inline bool lbr_is_reset_in_cstate(void *ctx)
> @@ -576,7 +576,7 @@ static void intel_pmu_arch_lbr_xsaves(vo
>   {
>   	struct x86_perf_task_context_arch_lbr_xsave *task_ctx = ctx;
>   
> -	copy_independent_supervisor_to_kernel(&task_ctx->xsave, XFEATURE_MASK_LBR);
> +	xsaves_to_kernel(&task_ctx->xsave, XFEATURE_MASK_LBR);
>   }
>   
>   static void __intel_pmu_lbr_save(void *ctx)
> @@ -992,7 +992,7 @@ static void intel_pmu_arch_lbr_read_xsav
>   		intel_pmu_store_lbr(cpuc, NULL);
>   		return;
>   	}
> -	copy_independent_supervisor_to_kernel(&xsave->xsave, XFEATURE_MASK_LBR);
> +	xsaves_to_kernel(&xsave->xsave, XFEATURE_MASK_LBR);
>   
>   	intel_pmu_store_lbr(cpuc, xsave->lbr.entries);
>   }

I tested the LBR Xsave feature on a Alder Lake machine. It looks good.

However, when I did other CPU hotplug test, it gave me an Oops.

$ sudo bash -c 'echo 0 > /sys/devices/system/cpu/cpu1/online'
$ sudo bash -c 'echo 1 > /sys/devices/system/cpu/cpu1/online'

[  108.912963] IRQ 132: no longer affine to CPU1
[  108.913010] IRQ148: set affinity failed(-22).
[  108.913038] IRQ149: set affinity failed(-22).
[  108.913050] IRQ150: set affinity failed(-22).
[  108.917436] smpboot: CPU 1 is now offline
[  111.191655] x86: Booting SMP configuration:
[  111.191661] smpboot: Booting Node 0 Processor 1 APIC 0x1
[  111.200452] BUG: unable to handle page fault for address: 
ffffffff996a96a0
[  111.207312] #PF: supervisor write access in kernel mode
[  111.207325] #PF: error_code(0x0003) - permissions violation
[  111.207335] PGD 40a02b067 P4D 40a02b067 PUD 40a02c063 PMD 106bdb063 
PTE 8000000409ea9161
[  111.218116] Oops: 0003 [#1] PREEMPT SMP NOPTI
[  111.218137] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 
5.13.0-rc5-perf+ #123
[  111.218156] RIP: 0010:fpu__init_cpu_xstate+0x3e/0x130
[  111.218184] Code: 00 00 00 48 8b 05 02 08 66 01 48 85 c0 0f 84 a7 00 
00 00 55 48 89 c6 4889 e5 53 81 e6 00 01 00 00 0f 85 b5 00 00 00 80 e4 
fe <48> 89 05 db 07 66 01 9c 58 0f 1f 4400 00 48 89 c3 fa 66 0f 1f 44
[  111.218195] RSP: 0000:ffffa466401abec0 EFLAGS: 00010002
[  111.273756] RAX: 0000000000000207 RBX: 0000000000000008 RCX: 
0000000000000000
[  111.273764] RDX: 0000000000310800 RSI: 0000000000000000 RDI: 
0000000080050033
[  111.273772] RBP: ffffa466401abec8 R08: 00000000fffffe00 R09: 
ffff98729f686078
[  111.273778] R10: ffffffff99826000 R11: ffffa466401abdde R12: 
0000000000000001
[  111.273783] R13: 0000000000000000 R14: ffff986f00c20000 R15: 
000000000000b000
[  111.273789] FS:  0000000000000000(0000) GS:ffff98729f680000(0000) 
knlGS:0000000000000000
[  111.273798] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  111.273805] CR2: ffffffff996a96a0 CR3: 000000040a026001 CR4: 
0000000000330ea0
[  111.273813] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 
0000000000000000
[  111.273817] DR3: 0000000000000000 DR6: 00000000ffff07f0 DR7: 
0000000000000400
[  111.273834] invalid opcode: 0000 [#2] PREEMPT SMP NOPTI
[  111.273843] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 
5.13.0-rc5-perf+ #123

Thanks,
Kan