Re: [PATCH 00/10] vhost: stick to -errno error return convention

[Roman Kagan <rvkagan@yandex-team.ru>]

On Sun, Nov 28, 2021 at 04:47:20PM -0500, Michael S. Tsirkin wrote:
> On Thu, Nov 11, 2021 at 06:33:44PM +0300, Roman Kagan wrote:
> > Error propagation between the generic vhost code and the specific backends is
> > not quite consistent: some places follow "return -1 and set errno" convention,
> > while others assume "return negated errno".  Furthermore, not enough care is
> > taken not to clobber errno.
> > 
> > As a result, on certain code paths the errno resulting from a failure may get
> > overridden by another function call, and then that zero errno inidicating
> > success is propagated up the stack, leading to failures being lost.  In
> > particular, we've seen errors in the communication with a vhost-user-blk slave
> > not trigger an immediate connection drop and reconnection, leaving it in a
> > broken state.
> > 
> > Rework error propagation to always return negated errno on errors and
> > correctly pass it up the stack.
> 
> Hi Roman,
> if there are bugfixes here I'll be happy to take them right now.
> The wholesale rework seems inappropriate for 6.2, I'll be
> happy to tag it for after 6.2. Pls ping me aftre release to help
> make sure it's not lost.

All these patches are bugfixes in one way or another.  That said, none
of the problems being addressed are recent regressions.  OTOH the
patches introduce non-zero churn and change behavior on some error
paths, so I'd suggest to postpone the whole series till after 6.2 is
out.

Thanks,
Roman.