Re: [PATCH] loop: don't print warnings if the underlying filesystem doesn't support discard

[Christoph Hellwig <hch@lst.de>]

On Thu, Sep 23, 2021 at 03:48:27PM -0400, Mikulas Patocka wrote:
> Hi
> 
> When running the lvm testsuite, we get a lot of warnings 
> "blk_update_request: operation not supported error, dev loop0, sector 0 op 
> 0x9:(WRITE_ZEROES) flags 0x800800 phys_seg 0 prio class 0". The lvm 
> testsuite puts the loop device on tmpfs and the reason for the warning is 
> that tmpfs supports fallocate, but doesn't support FALLOC_FL_ZERO_RANGE.
> 
> I've created this patch to silence the warnings.
> 
> Mikulas
> 
> 
> 
> From: Mikulas Patocka <mpatocka@redhat.com>
> 
> The loop driver checks for the fallocate method and if it is present, it
> assumes that the filesystem can do FALLOC_FL_ZERO_RANGE and
> FALLOC_FL_PUNCH_HOLE requests. However, some filesystems (such as fat, or
> tmpfs) have the fallocate method, but lack the capability to do
> FALLOC_FL_ZERO_RANGE and/or FALLOC_FL_PUNCH_HOLE.
> 
> This results in syslog warnings "blk_update_request: operation not
> supported error, dev loop0, sector 0 op 0x9:(WRITE_ZEROES) flags 0x800800
> phys_seg 0 prio class 0"
> 
> This patch sets RQF_QUIET to silence the warnings.

Doesn't this just paper over the problem?  I think we need an
unsigned int with flag in the file_operations for the supported
operations for this kind of use.