Re: next: kernel BUG at drivers/virtio/virtio.c:122!

["Michael S. Tsirkin" <mst@redhat.com>]

On Tue, Dec 20, 2022 at 02:17:45PM +0100, Marek Szyprowski wrote:
> On 20.12.2022 11:54, Michael S. Tsirkin wrote:
> > On Tue, Dec 20, 2022 at 06:27:59PM +0800, Xuan Zhuo wrote:
> >> On Tue, 20 Dec 2022 14:51:54 +0530, Naresh Kamboju <naresh.kamboju@linaro.org> wrote:
> >>> The qemu-x86_64 and qemu-arm64 boot failed with Linux next-20221220 tag.
> >>> It is always reproducible with gcc-11/ gcc-12 and clang tool chains.
> >>>
> >>> Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
> >>>
> >>> <6>[    0.000000] Booting Linux on physical CPU 0x0000000000 [0x000f0510]
> >>> <5>[    0.000000] Linux version 6.1.0-next-20221220 (tuxmake@tuxmake)
> >>> (Debian clang version 16.0.0
> >>> (++20221218072110+9c1b82599dac-1~exp1~20221218072217.501), Debian LLD
> >>> 16.0.0) #1 SMP PREEMPT @1671498176
> >>> <5>[    0.000000] random: crng init done
> >>> <6>[    0.000000] Machine model: linux,dummy-virt
> >>> ....
> >>> <6>[    3.571098] loop: module loaded
> >>> <6>[    3.573410] virtio_blk virtio0: 1/0/0 default/read/poll queues
> >>> <5>[    3.578816] virtio_blk virtio0: [vda] 2816420 512-byte logical
> >>> blocks (1.44 GB/1.34 GiB)
> >>> <4>[    3.581234] ------------[ cut here ]------------
> >>> <2>[    3.581595] kernel BUG at drivers/virtio/virtio.c:122!
> >>> <0>[    3.582906] Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP
> >>> <4>[    3.583672] Modules linked in:
> >>> <4>[    3.584905] CPU: 0 PID: 1 Comm: swapper/0 Not tainted
> >>> 6.1.0-next-20221220 #1
> >>> <4>[    3.585801] Hardware name: linux,dummy-virt (DT)
> >>> <4>[    3.586591] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT
> >>> -SSBS BTYPE=--)
> >>> <4>[    3.587349] pc : virtio_check_driver_offered_feature+0x60/0x6c
> >>> <4>[    3.588613] lr : virtblk_probe+0x7cc/0x8f0
> >>> <4>[    3.588915] sp : ffff80000802b840
> >>> <4>[    3.589314] x29: ffff80000802b850 x28: 0000000000000000 x27:
> >>> 0000000000000001
> >>> <4>[    3.590509] x26: 0000000000000001 x25: 0000000000000200 x24:
> >>> ffff0000c1699b00
> >>> <4>[    3.590832] x23: 00000000000000fe x22: ffff0000c0313c30 x21:
> >>> ffff0000c0313c00
> >>> <4>[    3.591568] x20: ffff0000c02f4080 x19: 0000000000000000 x18:
> >>> ffffffffffffffff
> >>> <4>[    3.592056] x17: 00000000ffffffea x16: 00000000fffffffe x15:
> >>> 0000000000000004
> >>> <4>[    3.592632] x14: 0000000000000fff x13: ffff800008028000 x12:
> >>> ffff80000802c000
> >>> <4>[    3.593157] x11: 000000000000000f x10: 000000000000000f x9 :
> >>> ffffae66eee314d8
> >>> <4>[    3.594048] x8 : 0000000000000000 x7 : 3631383837352e33 x6 :
> >>> 202020205b3e353c
> >>> <4>[    3.594429] x5 : 0000000000000001 x4 : 0000000000000000 x3 :
> >>> ffff80000802b780
> >>> <4>[    3.594956] x2 : ffff80000802b6f8 x1 : 0000000000000011 x0 :
> >>> ffff0000c02f4080
> >>> <4>[    3.595811] Call trace:
> >>> <4>[    3.596120]  virtio_check_driver_offered_feature+0x60/0x6c
> >>> <4>[    3.596632]  virtio_dev_probe+0x274/0x320
> >>> <4>[    3.597064]  really_probe+0x178/0x418
> >>> <4>[    3.597547]  __driver_probe_device+0x120/0x188
> >>> <4>[    3.597820]  driver_probe_device+0x48/0x22c
> >>> <4>[    3.598189]  __driver_attach+0xf8/0x250
> >>> <4>[    3.598656]  bus_for_each_dev+0x8c/0xd8
> >>> <4>[    3.599151]  driver_attach+0x30/0x3c
> >>> <4>[    3.599321]  bus_add_driver+0x11c/0x22c
> >>> <4>[    3.599684]  driver_register+0x84/0x120
> >>> <4>[    3.600175]  register_virtio_driver+0x38/0x48
> >>> <4>[    3.600575]  virtio_blk_init+0x6c/0xb4
> >>> <4>[    3.600759]  do_one_initcall+0xe0/0x2f8
> >>> <4>[    3.601197]  do_initcall_level+0xa8/0x150
> >>> <4>[    3.601703]  do_initcalls+0x60/0xa0
> >>> <4>[    3.601892]  do_basic_setup+0x28/0x34
> >>> <4>[    3.602229]  kernel_init_freeable+0x100/0x178
> >>> <4>[    3.602706]  kernel_init+0x2c/0x1b4
> >>> <4>[    3.603162]  ret_from_fork+0x10/0x20
> >>> <0>[    3.604398] Code: 540000a0 91001129 f1000508 54ffff61 (d4210000)
> >>> <4>[    3.606108] ---[ end trace 0000000000000000 ]---
> >>> <6>[    3.606714] note: swapper/0[1] exited with preempt_count 1
> >>> <0>[    3.608155] Kernel panic - not syncing: Attempted to kill init!
> >>> exitcode=0x0000000b
> >>> <2>[    3.609173] SMP: stopping secondary CPUs
> >>> <0>[    3.610506] Kernel Offset: 0x2e66e4800000 from 0xffff800008000000
> >>> <0>[    3.610970] PHYS_OFFSET: 0x40000000
> >>> <0>[    3.611428] CPU features: 0x00000,003a612f,cd22773f
> >>> <0>[    3.612277] Memory Limit: none
> >>> <0>[    3.613192] ---[ end Kernel panic - not syncing: Attempted to
> >>> kill init! exitcode=0x0000000b ]---
> >>>
> >>> Ref Links,
> >>> https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20221220/testrun/13764099/suite/log-parser-test/test/check-kernel-bug/details/
> >>> https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20221220/testrun/13766814/suite/boot/test/clang-nightly-lkftconfig/details/
> >>> https://storage.tuxsuite.com/public/linaro/lkft/builds/2J9l3qxSBmPWVa1S2uWn0Xuwl1J/config
> >>>
> >>
> >> I think there is a bug, VIRTIO_BLK_F_ZONED depends the CONFIG_BLK_DEV_ZONE, but
> >> checking its code in virtblk_probe does not depend on this config.
> >>
> >> This option in your config file is indeed not opened.
> >>
> >>
> >> static int virtblk_probe(struct virtio_device *vdev){
> >> .....
> >> 	virtblk_update_capacity(vblk, false);
> >> 	virtio_device_ready(vdev);
> >>
> >> 	if (virtio_has_feature(vdev, VIRTIO_BLK_F_ZONED)) {
> >> 		err = virtblk_probe_zoned_device(vdev, vblk, q);
> >> 		if (err)
> >> 			goto out_cleanup_disk;
> >> 	}
> >>
> >> 	dev_info(&vdev->dev, "blk config size: %zu\n",
> >> 		sizeof(struct virtio_blk_config));
> >>
> >> ....
> >>
> >> static unsigned int features[] = {
> >> 	VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX, VIRTIO_BLK_F_GEOMETRY,
> >> 	VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
> >> 	VIRTIO_BLK_F_FLUSH, VIRTIO_BLK_F_TOPOLOGY, VIRTIO_BLK_F_CONFIG_WCE,
> >> 	VIRTIO_BLK_F_MQ, VIRTIO_BLK_F_DISCARD, VIRTIO_BLK_F_WRITE_ZEROES,
> >> 	VIRTIO_BLK_F_SECURE_ERASE, VIRTIO_BLK_F_LIFETIME,
> >> #ifdef CONFIG_BLK_DEV_ZONED
> >> 	VIRTIO_BLK_F_ZONED,
> >> #endif /* CONFIG_BLK_DEV_ZONED */
> >> }
> >>
> >>
> >> Thanks.
> > Oh you are right, I wanted to test it so of course I enabled it.
> >
> > Testing this fixup now:
> >
> > commit f9503d0fd839b59e0f48b00c20635437c41420ed
> > Author: Michael S. Tsirkin <mst@redhat.com>
> > Date:   Tue Dec 20 05:53:41 2022 -0500
> >
> >      fixup! virtio-blk: add support for zoned block devices
> 
> I also confirm that this patch fixes the issue with broken virtio block 
> device observed on QEMU arm/arm64 'virt' machines and linux 
> next-20221220. Thanks for the fix!
> 
> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>

Merged the fix already so I'm not sure I can add this tag.
If I rebase before pull request I will, thanks!

> > diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
> > index ff49052e26f7..d8cdaeb51d1c 100644
> > --- a/drivers/block/virtio_blk.c
> > +++ b/drivers/block/virtio_blk.c
> > @@ -775,6 +775,10 @@ static int virtblk_probe_zoned_device(struct virtio_device *vdev,
> >   	return ret;
> >   }
> >   
> > +static inline bool virtblk_has_zoned_feature(struct virtio_device *vdev)
> > +{
> > +	return virtio_has_feature(vdev, VIRTIO_BLK_F_ZONED);
> > +}
> >   #else
> >   
> >   /*
> > @@ -790,6 +794,11 @@ static inline int virtblk_probe_zoned_device(struct virtio_device *vdev,
> >   {
> >   	return -EOPNOTSUPP;
> >   }
> > +
> > +static inline bool virtblk_has_zoned_feature(struct virtio_device *vdev)
> > +{
> > +	return false;
> > +}
> >   #endif /* CONFIG_BLK_DEV_ZONED */
> >   
> >   /* return id (s/n) string for *disk to *id_str
> > @@ -1576,11 +1585,11 @@ static int virtblk_probe(struct virtio_device *vdev)
> >   		else
> >   			q->limits.discard_granularity = blk_size;
> >   	}
> >   
> >   	virtblk_update_capacity(vblk, false);
> >   	virtio_device_ready(vdev);
> >   
> > -	if (virtio_has_feature(vdev, VIRTIO_BLK_F_ZONED)) {
> > +	if (virtblk_has_zoned_feature(vdev)) {
> >   		err = virtblk_probe_zoned_device(vdev, vblk, q);
> >   		if (err)
> >   			goto out_cleanup_disk;
> 
> Best regards
> -- 
> Marek Szyprowski, PhD
> Samsung R&D Institute Poland