debian/tests/corosync: gfs2_jadd fails with ENOTTY for i386 package on amd64 kernel

[Mauricio Faria de Oliveira]

[Impact]

 * The gfs2_jadd command (used in debian/tests/corosync)
   fails with ENOTTY ("Inappropriate ioctl for device")
   when the i386 packages are run on an amd64 kernel.

 * The i386 kernel is gone, only amd64 is available.

 * The fix introduces the compat ioctl()s for gfs2,
   which landed in v5.5-rc1 thus not in Eoan/Focal.

[Test Case]

 * The test steps are detailed in comment #2
   based on gfs2-utils:debian/tests/corosync,
   modified to be simpler and to fail either
   with ENOTTY (failure) or ENOSPC (success).

[Regression Potential]

 * The previous surface for this problem is
   likely low, as i386 arch installs (which
   is not as popular/widely used as amd64
   nowadays) probably used the i386 kernel.

 * The amd64-only kernel version released
   is 5.3.0-17.18 dated 2019-10-10; gives
   ~3 month window to mid January 2020.

 * So, behavior changes could theoretically
   be seen on i386 installs with the newer
   kernel which is amd64-only, where gfs2-utils
   failed with ENOTTY and now is working fine
   or failing differently (e.g. see test-case).

[Original Description]

The gfs2_jadd command (used in debian/tests/corosync)
fails with ENOTTY ("Inappropriate ioctl for device")
when the i386 packages are run on an amd64 kernel.

The i386 kernel has gone away, only amd64 is available:

- The i386 kernel build is dropped as of 5.3.0-15.16:
  * Drop i386 build for 19.10 (LP: #1845714)

- Only linux-libc-dev:i386 is still built as of 5.3.0-17.18
  * Re-enable linux-libc-dev build on i386 (LP: #1846508)

The last passing autopkgtests for gfs2-utils:i386 are on 5.3.0-13.14 built for i386.
The same kernel version built for amd64 _fails_.
(so it's potentially user/kernel interface, or kernel-only.)

- i386/i686)

# uname -rvm
5.3.0-13-generic #14-Ubuntu SMP Tue Sep 24 02:43:47 UTC 2019 i686

root@sf256930ei:~# gfs2_jadd -j 1 /mnt
Filesystem: /mnt
Old journals: 2
New journals: 3

- amd64/x86_64)

# uname -rvm
5.3.0-13-generic #14-Ubuntu SMP Tue Sep 24 02:46:08 UTC 2019 x86_64

# gfs2_jadd -j 1 /mnt
GETFLAGS: Inappropriate ioctl for device

[Mauricio Faria de Oliveira (mfo)]

This issue should be fixed with linux commit 8d0980704842 ("gfs2: add compat_ioctl support") [1] introduced in v5.5-rc1 which adds handling for ioctl FS_IOC32_GETFLAGS and FS_IOC32_SETFLAGS.
Kicking a test build.

$ git describe --contains 8d0980704842e
v5.5-rc1~101^2~14

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=8d0980704842e

[Mauricio Faria de Oliveira (mfo)]

Test Environment:
---

Eoan/Focal amd64 kernel on i386 dpkg arch.

$ dpkg-architecture | grep DEB_HOST_ARCH=
DEB_HOST_ARCH=i386

$ uname -m
x86_64

Test Steps:
---

apt install -y gfs2-utils corosync dlm-controld linux-modules-extra-$(uname -r)

corosync-quorumtool -e 1
systemctl restart dlm.service

DISK=disk.img
dd if=/dev/zero of=$DISK bs=1M count=128 2>&1
LOOP=$(losetup --find --show $DISK)

mkfs.gfs2 -p lock_dlm -t debian:test -J 32 -j 2 -O $LOOP
mount $LOOP /mnt

gfs2_jadd -j 1 /mnt

Note:
The 'gfs2_jadd' command either fails with ENOTTY upfront or goes on.
For a simple test let's make it fail with ENOSPC instead.

[Mauricio Faria de Oliveira (mfo)]

Test on Eoan:
---

Before:

# uname -rv
5.3.0-27-generic #29-Ubuntu SMP Thu Jan 9 01:36:10 UTC 2020

gfs2_jadd -j 1 /mnt
GETFLAGS: Inappropriate ioctl for device

After:

# uname -rv
5.3.0-27-generic #29+gfs2compatioctl SMP Wed Jan 15 13:56:12 UTC 2020

# gfs2_jadd -j 1 /mnt
add_j: No space left on device

[Mauricio Faria de Oliveira (mfo)]

Test on Focal:
---

Before:

# uname -rv
5.4.0-9-generic #12-Ubuntu SMP Mon Dec 16 22:34:19 UTC 2019

# gfs2_jadd -j 1 /mnt
GETFLAGS: Inappropriate ioctl for device

After:

# uname -rv
5.4.0-11-generic #14+gfs2compatioctl.1 SMP Wed Jan 15 13:40:46 UTC 2020

# gfs2_jadd -j 1 /mnt
add_j: No space left on device

[Mauricio Faria de Oliveira (mfo)]

[E/F][PATCH] gfs2: add compat_ioctl support
https://lists.ubuntu.com/archives/kernel-team/2020-January/106913.html

[Rafael David Tinoco (rafaeldtinoco)]

Very nice! Thanks Mauricio!

Should we flag this bug to linux instead of gfs2-utils ?

[Mauricio Faria de Oliveira (mfo)]

Rafael, oh indeed; I missed updating the affected packages. Thanks for catching that! Done.

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-eoan' to 'verification-done-eoan'. If the problem still exists, change the tag 'verification-needed-eoan' to 'verification-failed-eoan'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

[Mauricio Faria de Oliveira (mfo)]

Verification done on Eoan.
The gfs2_jadd i386 command on amd64 kernel no longer fails with ENOTTY upfront.

# lsb_release -cs
eoan

# dpkg --print-architecture
i386

# uname -m
x86_64

# uname -rv
5.3.0-40-generic #32-Ubuntu SMP Fri Jan 31 20:24:34 UTC 2020

# apt-cache madison linux-image-5.3.0-40-generic
linux-image-5.3.0-40-generic:amd64 | 5.3.0-40.32 | http://archive.ubuntu.com/ubuntu eoan-proposed/main amd64 Packages

# apt install -y gfs2-utils corosync dlm-controld linux-modules-extra-$(uname -r)

# corosync-quorumtool -e 1
# systemctl restart dlm.service

# DISK=disk.img
# dd if=/dev/zero of=$DISK bs=1M count=128 2>&1
# LOOP=$(losetup --find --show $DISK)

# mkfs.gfs2 -p lock_dlm -t debian:test -J 32 -j 2 -O $LOOP
# mount $LOOP /mnt

# gfs2_jadd -j 1 /mnt
add_j: No space left on device

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 5.3.0-40.32

---------------
linux (5.3.0-40.32) eoan; urgency=medium

  * eoan/linux: 5.3.0-40.32 -proposed tracker (LP: #1861214)

  * No sof soundcard for 'ASoC: CODEC DAI intel-hdmi-hifi1 not registered' after
    modprobe sof (LP: #1860248)
    - ASoC: SOF: Intel: fix HDA codec driver probe with multiple controllers

  * ocfs2-tools is causing kernel panics in Ubuntu Focal (Ubuntu-5.4.0-9.12)
    (LP: #1852122)
    - ocfs2: fix the crash due to call ocfs2_get_dlm_debug once less

  * QAT drivers for C3XXX and C62X not included as modules (LP: #1845959)
    - [Config] CRYPTO_DEV_QAT_C3XXX=m, CRYPTO_DEV_QAT_C62X=m and
      CRYPTO_DEV_QAT_DH895xCC=m

  * Eoan update: upstream stable patchset 2020-01-24 (LP: #1860816)
    - scsi: lpfc: Fix discovery failures when target device connectivity bounces
    - scsi: mpt3sas: Fix clear pending bit in ioctl status
    - scsi: lpfc: Fix locking on mailbox command completion
    - Input: atmel_mxt_ts - disable IRQ across suspend
    - f2fs: fix to update time in lazytime mode
    - iommu: rockchip: Free domain on .domain_free
    - iommu/tegra-smmu: Fix page tables in > 4 GiB memory
    - dmaengine: xilinx_dma: Clear desc_pendingcount in xilinx_dma_reset
    - scsi: target: compare full CHAP_A Algorithm strings
    - scsi: lpfc: Fix SLI3 hba in loop mode not discovering devices
    - scsi: csiostor: Don't enable IRQs too early
    - scsi: hisi_sas: Replace in_softirq() check in hisi_sas_task_exec()
    - powerpc/pseries: Mark accumulate_stolen_time() as notrace
    - powerpc/pseries: Don't fail hash page table insert for bolted mapping
    - powerpc/tools: Don't quote $objdump in scripts
    - dma-debug: add a schedule point in debug_dma_dump_mappings()
    - leds: lm3692x: Handle failure to probe the regulator
    - clocksource/drivers/asm9260: Add a check for of_clk_get
    - clocksource/drivers/timer-of: Use unique device name instead of timer
    - powerpc/security/book3s64: Report L1TF status in sysfs
    - powerpc/book3s64/hash: Add cond_resched to avoid soft lockup warning
    - ext4: update direct I/O read lock pattern for IOCB_NOWAIT
    - ext4: iomap that extends beyond EOF should be marked dirty
    - jbd2: Fix statistics for the number of logged blocks
    - scsi: tracing: Fix handling of TRANSFER LENGTH == 0 for READ(6) and WRITE(6)
    - scsi: lpfc: Fix duplicate unreg_rpi error in port offline flow
    - f2fs: fix to update dir's i_pino during cross_rename
    - clk: qcom: Allow constant ratio freq tables for rcg
    - clk: clk-gpio: propagate rate change to parent
    - irqchip/irq-bcm7038-l1: Enable parent IRQ if necessary
    - irqchip: ingenic: Error out if IRQ domain creation failed
    - fs/quota: handle overflows of sysctl fs.quota.* and report as unsigned long
    - scsi: lpfc: fix: Coverity: lpfc_cmpl_els_rsp(): Null pointer dereferences
    - PCI: rpaphp: Fix up pointer to first drc-info entry
    - scsi: ufs: fix potential bug which ends in system hang
    - powerpc/pseries/cmm: Implement release() function for sysfs device
    - PCI: rpaphp: Don't rely on firmware feature to imply drc-info support
    - PCI: rpaphp: Annotate and corr...

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 5.4.0-18.22

---------------
linux (5.4.0-18.22) focal; urgency=medium

  * focal/linux: 5.4.0-18.22 -proposed tracker (LP: #1866488)

  * Packaging resync (LP: #1786013)
    - [Packaging] resync getabis
    - [Packaging] update helper scripts

  * Add sysfs attribute to show remapped NVMe (LP: #1863621)
    - SAUCE: ata: ahci: Add sysfs attribute to show remapped NVMe device count

  * [20.04 FEAT] Compression improvements in Linux kernel (LP: #1830208)
    - lib/zlib: add s390 hardware support for kernel zlib_deflate
    - s390/boot: rename HEAP_SIZE due to name collision
    - lib/zlib: add s390 hardware support for kernel zlib_inflate
    - s390/boot: add dfltcc= kernel command line parameter
    - lib/zlib: add zlib_deflate_dfltcc_enabled() function
    - btrfs: use larger zlib buffer for s390 hardware compression
    - [Config] Introducing s390x specific kernel config option CONFIG_ZLIB_DFLTCC

  * [UBUNTU 20.04] s390x/pci: increase CONFIG_PCI_NR_FUNCTIONS to 512 in kernel
    config (LP: #1866056)
    - [Config] Increase CONFIG_PCI_NR_FUNCTIONS from 64 to 512 starting with focal
      on s390x

  * CONFIG_IP_MROUTE_MULTIPLE_TABLES is not set (LP: #1865332)
    - [Config] CONFIG_IP_MROUTE_MULTIPLE_TABLES=y

  * Dell XPS 13 9300 Intel 1650S wifi [34f0:1651] fails to load firmware
    (LP: #1865962)
    - iwlwifi: remove IWL_DEVICE_22560/IWL_DEVICE_FAMILY_22560
    - iwlwifi: 22000: fix some indentation
    - iwlwifi: pcie: rx: use rxq queue_size instead of constant
    - iwlwifi: allocate more receive buffers for HE devices
    - iwlwifi: remove some outdated iwl22000 configurations
    - iwlwifi: assume the driver_data is a trans_cfg, but allow full cfg

  * [FOCAL][REGRESSION] Intel Gen 9 brightness cannot be controlled
    (LP: #1861521)
    - Revert "USUNTU: SAUCE: drm/i915: Force DPCD backlight mode on Dell Precision
      4K sku"
    - Revert "UBUNTU: SAUCE: drm/i915: Force DPCD backlight mode on X1 Extreme 2nd
      Gen 4K AMOLED panel"
    - SAUCE: drm/dp: Introduce EDID-based quirks
    - SAUCE: drm/i915: Force DPCD backlight mode on X1 Extreme 2nd Gen 4K AMOLED
      panel
    - SAUCE: drm/i915: Force DPCD backlight mode for some Dell CML 2020 panels

  * [20.04 FEAT] Enable proper kprobes on ftrace support (LP: #1865858)
    - s390/ftrace: save traced function caller
    - s390: support KPROBES_ON_FTRACE

  * alsa/sof: load different firmware on different platforms (LP: #1857409)
    - ASoC: SOF: Intel: hda: use fallback for firmware name
    - ASoC: Intel: acpi-match: split CNL tables in three
    - ASoC: SOF: Intel: Fix CFL and CML FW nocodec binary names.

  * [UBUNTU 20.04] Enable CONFIG_NET_SWITCHDEV in kernel config for s390x
    starting with focal (LP: #1865452)
    - [Config] Enable CONFIG_NET_SWITCHDEV in kernel config for s390x starting
      with focal

  * Focal update: v5.4.24 upstream stable release (LP: #1866333)
    - io_uring: grab ->fs as part of async offload
    - EDAC: skx_common: downgrade message importance on missing PCI device
    - net: dsa: b53: Ensure the default VID is untagged
    - net: fib_rules: Correctly set table field when table number exceeds 8 bit...