Re: [BUG BISECT] NFS root failure (NULL pointer)

From: Krzysztof Kozlowski <krzk@kernel.org>
To: dhowells@redhat.com, ebiggers@google.com,
	viro@zeniv.linux.org.uk, linux-fsdevel@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	"linux-samsung-soc@vger.kernel.org" 
	<linux-samsung-soc@vger.kernel.org>
Subject: Re: [BUG BISECT] NFS root failure (NULL pointer)
Date: Mon, 10 Sep 2018 08:44:15 +0200	[thread overview]
Message-ID: <CAJKOXPfc5O0c6HBtM=h2NUXwaiCWVgFC9eVeirLYynfUpU6d3Q@mail.gmail.com> (raw)
In-Reply-To: <CAJKOXPc49jh8RCDvn2Meci9iDcnnwGEHo6F7RuhkXpen=f_E8w@mail.gmail.com>

On Thu, 6 Sep 2018 at 09:10, Krzysztof Kozlowski <krzk@kernel.org> wrote:
>
> Hi,
>
> Today's next fails to mount NFS root under my ARM targets and fails to
> mount root from file image under QMU.
>
> [ 21.512866] Unable to handle kernel NULL pointer dereference at
> virtual address 00000000
> [ 21.695484] [<c03bcc04>] (nfs_fs_mount) from [<c02cf6a4>]
> (legacy_get_tree+0x34/0xec)
> [ 21.703225] [<c02cf6a4>] (legacy_get_tree) from [<c029442c>]
> (vfs_get_tree+0x64/0x180)
> [ 21.711119] [<c029442c>] (vfs_get_tree) from [<c02b9140>]
> (do_mount+0x21c/0x958)
> [ 21.718478] [<c02b9140>] (do_mount) from [<c02b9c38>] (ksys_mount+0x8c/0xbc)
> [ 21.725513] [<c02b9c38>] (ksys_mount) from [<c0101000>]
> (ret_fast_syscall+0x0/0x28)
>
> Full log from ARM (NFS root):
> https://krzk.eu/#/builders/25/builds/750/steps/12/logs/serial0
>
> The NFS root failure bisected to:
> bae551929c5433bd56ec4dcb97c7d4a50153d357 is the first bad commit
> commit bae551929c5433bd56ec4dcb97c7d4a50153d357
> Author: David Howells <dhowells@redhat.com>
> Date:   Tue Jul 10 21:43:37 2018 +0100

FYI, the bug is still present in linux-next. All boards fail to boot up.

Let me know if I can help anymore in debugging this.

Best regards,
Krzysztof

>
>     vfs: Separate changing mount flags full remount
>
>     Separate just the changing of mount flags (MS_REMOUNT|MS_BIND) from full
>     remount because the mount data will get parsed with the new fs_context
>     stuff prior to doing a remount - and this causes the syscall to fail under
>     some circumstances.
>
> The QEMU issue seems slightly different and I did not bisect it. The
> QEMU just cannot find rootfs:
> [    1.008052] Filesystem requires source device
> [    1.008513] VFS: Cannot open root device "mmcblk0" or
> unknown-block(179,0): error -2
> [    1.008790] Please append a correct "root=" boot option; here are
> the available partitions:
> [    1.009300] 0100           16384 ram0
> [    1.009337]  (driver?)
> [    1.009806] fe00            5120 vda
> [    1.009843]  driver: virtio_blk
> [    1.010296] b300          110592 mmcblk0
> [    1.010319]  driver: mmcblk
> [    1.010859] Kernel panic - not syncing: VFS: Unable to mount root
> fs on unknown-block(179,0)
> [    1.011580] ---[ end Kernel panic - not syncing: VFS: Unable to
> mount root fs on unknown-block(179,0) ]---
>
>
>
> Boards config:
> 1. Arch ARM Linux
> 2. exynos_defconfig
>   - Odroid HC1
>     ARMv7, octa-core (Cortex-A7+A15), Exynos5422 SoC
>     Systemd: v239
>   - Odroid U3
>     ARMv7, quad-core, Exynos4412 SoC
>     Systemd: v238
>   - Odroid XU
>     ARMv7, octa-core (Cortex-A7+A15) but only A15 working, Exynos5410 SoC
>     Systemd: v236
>   - Odroid XU3
>     ARMv7, octa-core (Cortex-A7+A15), Exynos5422 SoC
>     Systemd: v236
> 3. Custom VF50 defconfig
>   - Toradex Colibri VF50 on Iris board
>     ARMv7, UP, Cortext-A5, NXP VF500
>     Systemd: v232
> 4. All boards boot from TFTP with NFS root (NFSv4)
> 5. QEMU
>     ARMv7, vexpress-v2p-ca9, 128 MB RAM
>
>
>
> Best regards,
> Krzysztof