Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <0000000000007646bd05d7f81943@google.com> <Ygo0DBn/uZyKwtbt@hirez.programming.kicks-ass.net>
In-Reply-To: <Ygo0DBn/uZyKwtbt@hirez.programming.kicks-ass.net>
From:   Song Liu <song@kernel.org>
Date:   Mon, 14 Feb 2022 22:42:24 -0800
Message-ID: <CAPhsuW7w2yECSu_bGqw8mAkh-HBTi7sv1==kZOTyCRN3zXoDoQ@mail.gmail.com>
Subject: Re: [syzbot] kernel BUG in __text_poke
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     syzbot <syzbot+87f65c75f4a72db05445@syzkaller.appspotmail.com>,
        Borislav Petkov <bp@alien8.de>, dave.hansen@linux.intel.com,
        "H. Peter Anvin" <hpa@zytor.com>, jgross@suse.com,
        jpoimboe@redhat.com, open list <linux-kernel@vger.kernel.org>,
        Ingo Molnar <mingo@redhat.com>,
        Networking <netdev@vger.kernel.org>,
        syzkaller-bugs@googlegroups.com,
        Thomas Gleixner <tglx@linutronix.de>, X86 ML <x86@kernel.org>,
        Alexei Starovoitov <alexei.starovoitov@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Feb 14, 2022 at 2:51 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
>
> Song..
>
> On Mon, Feb 14, 2022 at 02:44:22AM -0800, syzbot wrote:
> > Hello,
> >
> > syzbot found the following issue on:
> >
> > HEAD commit:    f95f768f0af4 bpf, x86_64: Fail gracefully on bpf_jit_binar..
> > git tree:       bpf-next
> > console output: https://syzkaller.appspot.com/x/log.txt?x=13fb08c2700000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=c40b67275bfe2a58
> > dashboard link: https://syzkaller.appspot.com/bug?extid=87f65c75f4a72db05445
> > compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> >
> > Unfortunately, I don't have any reproducer for this issue yet.
> >
> > IMPORTANT: if you fix the issue, please add the following tag to the commit:
> > Reported-by: syzbot+87f65c75f4a72db05445@syzkaller.appspotmail.com
> >
> > ------------[ cut here ]------------
> > kernel BUG at arch/x86/kernel/alternative.c:989!
> > invalid opcode: 0000 [#1] PREEMPT SMP KASAN
> > CPU: 0 PID: 12993 Comm: syz-executor.1 Not tainted 5.16.0-syzkaller-11632-gf95f768f0af4 #0
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> > RIP: 0010:__text_poke+0x343/0x8c0 arch/x86/kernel/alternative.c:989
> > Code: c3 0f 86 2c fe ff ff 49 8d bc 24 00 10 00 00 e8 43 be 88 00 48 89 44 24 28 48 85 db 74 0c 48 83 7c 24 28 00 0f 85 1b fe ff ff <0f> 0b 48 b8 00 f0 ff ff ff ff 0f 00 49 21 c0 48 85 db 0f 85 c6 02
> > RSP: 0018:ffffc90005e6f7a8 EFLAGS: 00010246
> > RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> > RDX: ffff88807d1c8000 RSI: ffffffff81b3c443 RDI: 0000000000000003
> > RBP: 0000000000000080 R08: 0000000000000000 R09: ffffc90005e6f7bf
> > R10: ffffffff81b3c3e1 R11: 0000000000000001 R12: ffffffffa0010e00
> > R13: 0000000000000080 R14: 0000000000000e80 R15: 0000000000002000
> > FS:  00007fd60b1d8700(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
> > CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: 00007f9329383090 CR3: 000000007c3bb000 CR4: 00000000003506f0
> > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000600
> > Call Trace:
> >  <TASK>
> >  text_poke_copy+0x66/0xa0 arch/x86/kernel/alternative.c:1132
> >  bpf_arch_text_copy+0x21/0x40 arch/x86/net/bpf_jit_comp.c:2426
> >  bpf_jit_binary_pack_finalize+0x43/0x170 kernel/bpf/core.c:1094
> >  bpf_int_jit_compile+0x9d5/0x12f0 arch/x86/net/bpf_jit_comp.c:2383
> >  bpf_prog_select_runtime+0x4d4/0x8a0 kernel/bpf/core.c:2159
> >  bpf_prog_load+0xfe6/0x2250 kernel/bpf/syscall.c:2349
> >  __sys_bpf+0x68a/0x59a0 kernel/bpf/syscall.c:4640
> >  __do_sys_bpf kernel/bpf/syscall.c:4744 [inline]
> >  __se_sys_bpf kernel/bpf/syscall.c:4742 [inline]
> >  __x64_sys_bpf+0x75/0xb0 kernel/bpf/syscall.c:4742
> >  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> >  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
> >  entry_SYSCALL_64_after_hwframe+0x44/0xae
> > RIP: 0033:0x7fd60c863059

I think this is the same issue as [1], that the 2MB page somehow got freed
while still in use. I couldn't spot any bug with bpf_prog_pack allocate/free
logic. I haven't got luck reproducing it either. Will continue tomorrow.

[1] https://lore.kernel.org/bpf/00000000000073b3e805d7fed17e@google.com/T/

> > Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
> > RSP: 002b:00007fd60b1d8168 EFLAGS: 00000246 ORIG_RAX: 0000000000000141
> > RAX: ffffffffffffffda RBX: 00007fd60c975f60 RCX: 00007fd60c863059
> > RDX: 0000000000000048 RSI: 0000000020000200 RDI: 0000000000000005
> > RBP: 00007fd60c8bd08d R08: 0000000000000000 R09: 0000000000000000
> > R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
> > R13: 00007ffec77d82ef R14: 00007fd60b1d8300 R15: 0000000000022000
> >  </TASK>
> > Modules linked in:
> > ---[ end trace 0000000000000000 ]---
> > RIP: 0010:__text_poke+0x343/0x8c0 arch/x86/kernel/alternative.c:989
> > Code: c3 0f 86 2c fe ff ff 49 8d bc 24 00 10 00 00 e8 43 be 88 00 48 89 44 24 28 48 85 db 74 0c 48 83 7c 24 28 00 0f 85 1b fe ff ff <0f> 0b 48 b8 00 f0 ff ff ff ff 0f 00 49 21 c0 48 85 db 0f 85 c6 02
> > RSP: 0018:ffffc90005e6f7a8 EFLAGS: 00010246
> >
> > RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> > RDX: ffff88807d1c8000 RSI: ffffffff81b3c443 RDI: 0000000000000003
> > RBP: 0000000000000080 R08: 0000000000000000 R09: ffffc90005e6f7bf
> > R10: ffffffff81b3c3e1 R11: 0000000000000001 R12: ffffffffa0010e00
> >
> >
> > ---
> > 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@googlegroups.com.
> >
> > syzbot will keep track of this issue. See:
> > https://goo.gl/tpsmEJ#status for how to communicate with syzbot.