Re: [PATCH] netfs: Fix setting NETFS_RREQ_ALL_QUEUED to be after all subreqs queued

[Ihor Solodrai]

On 2/12/25 1:47 AM, David Howells wrote:
> Hi Ihor,
>
> Okay, the bug you're hitting appears to be a different one to the one I
> thought first.  Can you try the attached patch?  I managed to reproduce it
> with AFS by injecting a delay.
>
> [...]
>
> netfs: Fix setting NETFS_RREQ_ALL_QUEUED to be after all subreqs queued
>
> [...]

Hi David.

I tried this patch locally, and then on BPF CI. It fixes the 9p hanging issue.
A couple of platforms and toolchains are tested there:
https://github.com/kernel-patches/vmtest/actions/runs/13291034531

Tested-by: Ihor Solodrai <ihor.solodrai@xxxxxxxxx>

Note that on CI the "netfs: Fix a number of read-retry hangs" patch [1]
is *not* applied. Only this one.

Thank you!

[1] https://lore.kernel.org/v9fs/3173328.1738024385@xxxxxxxxxxxxxxxxxxxxxx/

> [...]