On Thu, Sep 22, 2022 at 5:12 PM Alberto Campinho Faria <afa...@redhat.com> wrote: > > On Thu, Sep 22, 2022 at 9:49 AM Paolo Bonzini <pbonz...@redhat.com> wrote: > > Callers of coroutine_fn must be coroutine_fn themselves, or the call > > must be within "if (qemu_in_coroutine())". Apply coroutine_fn to > > functions where this holds. > > > > Signed-off-by: Paolo Bonzini <pbonz...@redhat.com> > > --- > > block.c | 6 +++--- > > block/block-backend.c | 10 +++++----- > > block/io.c | 22 +++++++++++----------- > > 3 files changed, 19 insertions(+), 19 deletions(-) > > > > diff --git a/block.c b/block.c > > index bc85f46eed..5c34ada53f 100644 > > --- a/block.c > > +++ b/block.c > > @@ -631,9 +631,9 @@ static int64_t > > create_file_fallback_truncate(BlockBackend *blk, > > * Helper function for bdrv_create_file_fallback(): Zero the first > > * sector to remove any potentially pre-existing image header. > > */ > > -static int create_file_fallback_zero_first_sector(BlockBackend *blk, > > - int64_t current_size, > > - Error **errp) > > +static int coroutine_fn > > create_file_fallback_zero_first_sector(BlockBackend *blk, > > + int64_t > > current_size, > > + Error > > **errp) > > Why mark this coroutine_fn? Maybe the intention was to also replace > the call to blk_pwrite_zeroes() with blk_co_pwrite_zeroes()?
Because at the time I wrote the patch, blk_pwrite_zeroes() was a coroutine_fn. :) It is called from bdrv_co_create_opts_simple which is coroutine_fn and performs I/O, so it should be a coroutine_fn. I have a few more patches to not go through the generated_co_wrappers, but I was curious to see if we could automate those changes through your tool. Paolo > > Regardless: > > Reviewed-by: Alberto Faria <afa...@redhat.com> >
