mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Andrei Vagin <avagin@google.com>
To: "Chang S. Bae" <chang.seok.bae@intel.com>
Cc: Daniel Wade <danjwade95@gmail.com>,
	Thomas Gleixner <tglx@kernel.org>, Ingo Molnar <mingo@redhat.com>,
	Borislav Petkov <bp@alien8.de>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	x86@kernel.org, "H . Peter Anvin" <hpa@zytor.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] x86/fpu: Put non-enabled dynamic features into init state on sigreturn
Date: Thu, 19 Nov 2026 07:50:45 +0000	[thread overview]
Message-ID: <av6qVXvdezQgH7BW@google.com> (raw)
In-Reply-To: <av5TI7uMazFmWYid@google.com>

On Thu, Nov 19, 2026 at 01:38:43AM +0000, Andrei Vagin wrote:
> On Wed, Nov 18, 2026 at 12:19:02PM -0800, Chang S. Bae wrote:
> > On 11/17/2026 3:26 AM, Daniel Wade wrote:
> > > On sigreturn restore_fpregs_from_user() restores the features the signal
> > > frame carries and inits the rest, but both masks are limited to
> > > fpstate::user_xfeatures. If the task lost the registers earlier in the
> > > syscall, a dynamic feature it has not enabled yet -- XTILE_DATA before
> > > its first AMX use -- still holds the previous task state, and
> > 
> > How could that happen?
> 
> The code relies on this: while a task owns the FPU registers, a dynamic
> feature it has not enabled is in its init state.
> restore_fpregs_from_fpstate() maintains that by restoring with
> fpu_kernel_cfg.max_features (the comment there describes exactly this
> case), and fpu__clear_user_states() does the same on signal delivery.
> 
> The 64-bit sigreturn path doesn't. restore_fpregs_from_user() restores
> from the user buffer with RFBM and init_bv both limited to
> fpstate->user_xfeatures, then calls fpregs_mark_activate(). That clears
> TIF_NEED_FPU_LOAD, so the full restore on exit to user, which would have
> put XTILE_DATA into init state, never runs.
> 
> If the registers still belonged to current, that's fine. But the task
> can lose them inside rt_sigreturn: the first XRSTOR from the frame
> faults and fault_in_readable() sleeps, or the task is simply preempted.
> Switching out only saves the registers. It doesn't clear them, so the
> next AMX user leaves its tiles there:
> 
>   CPU0
>   A: rt_sigreturn()            A has AMX permission, XTILE_DATA not enabled
>        sleeps / preempted      A's state saved, TIF_NEED_FPU_LOAD set
>   B: runs, uses AMX            B's tiles are now in the registers
>   A: back in rt_sigreturn()
>        restore_fpregs_from_user()
>          xfd_update_state()    XFD[18] set: B's tiles hidden, not cleared
>          XRSTOR(frame, RFBM)   bit 18 not in RFBM or init_bv: untouched
>          fpregs_mark_activate()
>   A: exit to user              no reload, B's tiles still in the registers
> 
> 
> Here is the selftest to rerpoduce the issue:
> 
> Subject: [PATCH] selftests/x86/amx: Check that sigreturn does not expose stale
>  XTILEDATA
>

The next patch should be applied on top of the previous patch:

diff --git a/tools/testing/selftests/x86/amx.c b/tools/testing/selftests/x86/amx.c
index 31268193c4cd..17fc113eda5c 100644
--- a/tools/testing/selftests/x86/amx.c
+++ b/tools/testing/selftests/x86/amx.c
@@ -895,8 +895,13 @@ static void test_sigreturn_stale_tiledata(void)
 		attacker = fork();
 		if (attacker < 0)
 			fatal_error("fork");
-		if (attacker == 0)
-			_exit(stale_tiledata_attacker(cpu, helper_cpu));
+		if (attacker == 0) {
+			int ret = stale_tiledata_attacker(cpu, helper_cpu);
+
+			/* _exit() does not flush stdio buffers. */
+			fflush(stdout);
+			_exit(ret);
+		}
 
 		if (waitpid(attacker, &status, 0) != attacker)
 			fatal_error("waitpid");
diff --git a/tools/testing/selftests/x86/xstate.h b/tools/testing/selftests/x86/xstate.h
index eedf0cab7ccb..7229f3e5c6d4 100644
--- a/tools/testing/selftests/x86/xstate.h
+++ b/tools/testing/selftests/x86/xstate.h
@@ -93,7 +93,8 @@ static inline void xrstor(struct xsave_buffer *xbuf, uint64_t rfbm)
 	uint32_t rfbm_lo = rfbm;
 
 	asm volatile("xrstor (%%rdi)"
-		     : : "D" (xbuf), "a" (rfbm_lo), "d" (rfbm_hi));
+		     : : "D" (xbuf), "a" (rfbm_lo), "d" (rfbm_hi)
+		     : "memory");
 }
 
 static inline uint64_t xgetbv(uint32_t index)

--
I pushed all changed in https://github.com/avagin/linux-task-diag/commits/wip/fpu-sigreturn-init-xtiledata/.

Thanks,
Andrei

      reply	other threads:[~2026-11-19  7:50 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-17 11:26 Daniel Wade
2026-11-18 20:19 ` Chang S. Bae
2026-11-19  1:38   ` Andrei Vagin
2026-11-19  7:50     ` Andrei Vagin [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=av6qVXvdezQgH7BW@google.com \
    --to=avagin@google.com \
    --cc=bp@alien8.de \
    --cc=chang.seok.bae@intel.com \
    --cc=danjwade95@gmail.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=tglx@kernel.org \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®