Re: [PATCH] vfs: missing inode operation should return a consistent error code

[Christian Brauner]

On Tue, Jul 07, 2026 at 03:27:07PM +0100, Mark Brown wrote:
> On Thu, Jul 02, 2026 at 09:47:24AM +0200, Christian Brauner wrote:
> > On Sun, 31 May 2026 12:49:46 +0200, Jori Koolstra wrote:
> > > vfs: missing inode operation should return a consistent error code
> > 
> > I've stuffed this on a separate branch. Let's see whether we see any
> > fallout.
> 
> I'm seeing a failure in the chdir01 LTP test on arm64, everything seems
> to work fine apart from vfat:
> 
> tst_buffers.c:57: TINFO: Test is using guarded buffers
> L5023tst_tmpdir.c:316: TINFO: Using /tmp/LTP_chdkVKCfH as tmpdir (tmpfs filesystem)
> L5024tst_device.c:98: TINFO: Found free device 0 '/dev/loop0'
> L5025tst_test.c:2047: TINFO: LTP version: 20260130
> L5026tst_test.c:2050: TINFO: Tested kernel: 7.2.0-rc2-next-20260707 #1 SMP PREEMPT @1783429165 aarch64
> L5027tst_kconfig.c:88: TINFO: Parsing kernel config '/proc/config.gz'
> L5028tst_test.c:1875: TINFO: Overall timeout per run is 0h 00m 40s
> 
> ...
> 
> tst_test.c:1985: TINFO: === Testing on vfat ===
> L5158tst_test.c:1290: TINFO: Formatting /dev/loop0 with vfat opts='' extra opts=''
> L5159tst_test.c:1302: TINFO: Mounting /dev/loop0 to /tmp/LTP_chdkVKCfH/mntpoint fstyp=vfat flags=0
> L5160chdir01.c:85: TBROK: Cannot create symlinks: EOPNOTSUPP (95)
> 
> Looks like a changed the error code here has confused the test's feature
> detection.  How many non-LTP applications have the same detection
> algorithm is a separate question.
> 
> A full log may be visible at:
> 
>   https://lava.sirena.org.uk/scheduler/job/2954697#L5157

Dropped.