On Fri, Apr 24, 2026 at 01:30:45AM +0200, Alejandro Colomar via Mutt-dev wrote:
I would prefer not passing NONBLOCK at all.  Not failing and not adding
extra code.

I think the check for EINTR and res < length_requested are required anyway.

So I thought the cast comparing to -1 was worth the clarity trade off below.

But if in fact the code would be fine I'll convert to ssign_t.

You may also need to update the type of 'length_requested', to prevent
mixing signed and unsigned types, which will rightfully give warnings.

I don't think I want to do that. I'll take a closer look at the code, but I believe some parts rely on length_requested always being positive.

I think it may be easier to just remove the casts.

I'll play with ssize_t then.

Thank you.

--
Kevin J. McCarthy
GPG Fingerprint: 8975 A9B3 3AA3 7910 385C  5308 ADEF 7684 8031 6BDA

Attachment: signature.asc
Description: PGP signature

Reply via email to