Message-ID: <aQs0Bkrbnqinlqss@J2N7QTR9R3>
Date: Wed, 5 Nov 2025 11:24:54 +0000
From: Mark Rutland <mark.rutland@....com>
To: syzbot <syzbot+d4ab35af21e99d07ce67@...kaller.appspotmail.com>
Cc: catalin.marinas@....com, linux-arm-kernel@...ts.infradead.org,
	linux-kernel@...r.kernel.org, syzkaller-bugs@...glegroups.com,
	will@...nel.org
Subject: Re: [syzbot] [arm?] WARNING in fpsimd_restore_current_state

On Tue, Nov 04, 2025 at 03:35:30PM -0800, syzbot wrote:
> Hello,

Hi,

Initial analysis / holding-reply below.

> syzbot found the following issue on:
> 
> HEAD commit:    dcb6fa37fd7b Linux 6.18-rc3
> git tree:       git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-kernelci
> console output: https://syzkaller.appspot.com/x/log.txt?x=11a79704580000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=b8b659f0cab27b22
> dashboard link: https://syzkaller.appspot.com/bug?extid=d4ab35af21e99d07ce67
> compiler:       Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8
> userspace arch: arm64
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=115fdf34580000
> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=1441d258580000
> 
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/ae6bec0d0398/disk-dcb6fa37.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/dcc732da66c3/vmlinux-dcb6fa37.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/301d1bbdecc2/Image-dcb6fa37.gz.xz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+d4ab35af21e99d07ce67@...kaller.appspotmail.com
> 
> ------------[ cut here ]------------
> WARNING: CPU: 0 PID: 6720 at arch/arm64/kernel/fpsimd.c:370 task_fpsimd_load arch/arm64/kernel/fpsimd.c:370 [inline]
> WARNING: CPU: 0 PID: 6720 at arch/arm64/kernel/fpsimd.c:370 fpsimd_restore_current_state+0x4cc/0x708 arch/arm64/kernel/fpsimd.c:1746

This warning is due to the saved state being in SVE format unexpectedly (with
TIF_SVE clear and PSTATE.SM clear). Within task_fpsimd_load() we have:

	switch (current->thread.fp_type) {
	...
	case FP_STATE_SVE:
		if (!thread_sm_enabled(&current->thread))
			WARN_ON_ONCE(!test_and_set_thread_flag(TIF_SVE));

There's definitely some bug that allowed the task to get into a state that
should be forbidden.

I see that the repro has:

	ptrace$setregset(0x4205, r0, 0x405, &(0x7f0000000000)={0x0, 0x300})

... which looks to be a write to NT_ARM_SVE.

That suggests there's some path through sve_set_common() that leaves the task
in an inconsistent state.

>From a qucik look, it appears that's possible as a result of a partial SVE
write; I'll go check and see what we can do to fix that.

Mark.

> Modules linked in:
> CPU: 0 UID: 0 PID: 6720 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT 
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/03/2025
> pstate: 83400005 (Nzcv daif +PAN -UAO +TCO +DIT -SSBS BTYPE=--)
> pc : task_fpsimd_load arch/arm64/kernel/fpsimd.c:370 [inline]
> pc : fpsimd_restore_current_state+0x4cc/0x708 arch/arm64/kernel/fpsimd.c:1746
> lr : task_fpsimd_load arch/arm64/kernel/fpsimd.c:370 [inline]
> lr : fpsimd_restore_current_state+0x4cc/0x708 arch/arm64/kernel/fpsimd.c:1746
> sp : ffff8000a1047de0
> x29: ffff8000a1047de0 x28: ffff0000dcfd9ec0 x27: 0000000000000000
> x26: 0000000000000000 x25: 0000000000000008 x24: 0000000000000040
> x23: 00000000000020ff x22: dfff800000000000 x21: 1fffe0001b9fb3d8
> x20: 0000000000000000 x19: ffff0000dcfd9ec0 x18: 1fffe000337db690
> x17: ffff80008ae011f8 x16: ffff80008052ae04 x15: 0000000000000001
> x14: 1fffe0001b9fb3d8 x13: 0000000000000000 x12: 0000000000000000
> x11: ffff60001b9fb3d9 x10: 0000000000ff0100 x9 : 0000000000000000
> x8 : ffff0000dcfd9ec0 x7 : 0000000000000000 x6 : 0000000000000000
> x5 : 0000000000000001 x4 : 0000000000000008 x3 : ffff8000801f8344
> x2 : 0000000000000001 x1 : 0000000000000000 x0 : 0000000000000000
> Call trace:
>  task_fpsimd_load arch/arm64/kernel/fpsimd.c:370 [inline] (P)
>  fpsimd_restore_current_state+0x4cc/0x708 arch/arm64/kernel/fpsimd.c:1746 (P)
>  arch_exit_to_user_mode_work arch/arm64/include/asm/entry-common.h:25 [inline]
>  exit_to_user_mode_loop+0xe4/0x178 kernel/entry/common.c:46
>  exit_to_user_mode_prepare include/linux/irq-entry-common.h:225 [inline]
>  arm64_exit_to_user_mode arch/arm64/kernel/entry-common.c:103 [inline]
>  el0_svc+0x170/0x254 arch/arm64/kernel/entry-common.c:747
>  el0t_64_sync_handler+0x84/0x12c arch/arm64/kernel/entry-common.c:765
>  el0t_64_sync+0x198/0x19c arch/arm64/kernel/entry.S:596
> irq event stamp: 129
> hardirqs last  enabled at (127): [<ffff80008ae151bc>] __raw_spin_unlock_irq include/linux/spinlock_api_smp.h:159 [inline]
> hardirqs last  enabled at (127): [<ffff80008ae151bc>] _raw_spin_unlock_irq+0x30/0x80 kernel/locking/spinlock.c:202
> hardirqs last disabled at (129): [<ffff80008adecb80>] el1_brk64+0x20/0x54 arch/arm64/kernel/entry-common.c:434
> softirqs last  enabled at (8): [<ffff8000801f95fc>] local_bh_enable+0x10/0x34 include/linux/bottom_half.h:32
> softirqs last disabled at (128): [<ffff8000801f95c8>] local_bh_disable+0x10/0x34 include/linux/bottom_half.h:19
> ---[ end trace 0000000000000000 ]---
> 
> 
> ---
> This report is generated by a bot. It may contain errors.
> See https://goo.gl/tpsmEJ for more information about syzbot.
> syzbot engineers can be reached at syzkaller@...glegroups.com.
> 
> syzbot will keep track of this issue. See:
> https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
> 
> If the report is already addressed, let syzbot know by replying with:
> #syz fix: exact-commit-title
> 
> If you want syzbot to run the reproducer, reply with:
> #syz test: git://repo/address.git branch-or-commit-hash
> If you attach or paste a git patch, syzbot will apply it before testing.
> 
> If you want to overwrite report's subsystems, reply with:
> #syz set subsystems: new-subsystem
> (See the list of subsystem names on the web dashboard)
> 
> If the report is a duplicate of another one, reply with:
> #syz dup: exact-subject-of-another-report
> 
> If you want to undo deduplication, reply with:
> #syz undup
> 

Powered by blists - more mailing lists