Jay Satiro

Results 297 comments of Jay Satiro
trafficstars

I'm proposing #13522 to address this issue with a normalization check and cover all the file functions that we wrap.

> According to [testclutch](https://testclutch.curl.se/static/reports/summary.html), neat > the first quiche pytest failures started with https://github.com/curl/curl/actions/runs/8715610798, which was commit [5b94ece](https://github.com/curl/curl/commit/5b94eced85276bf2d3781fe32bb5e4a81cfd7145) - which seems totally innocuous and unrelated... I'm puzzled. Yeah that's the...

I really dislike making debug output dependent on the position of the -v option because it's confusing and not intuitive. IMO it's too late for -vv since that would be...

> > > CURL_RC set to a (full) file name to read by default instead of $HOME/.curlrc > > > I think you're overdoing it here. Why do we need...

It looks to me like Curl_resolver_kill is supposed to wait for threads to terminate and doesn't do that for the GetAddrInfoExW threads. https://github.com/curl/curl/blob/de7b3e89218467159a7af72d58cea8425946e97d/lib/asyn-thread.c#L746-L762 I am not sure how quick exit...

> I propose the simpler take: use `Retry-After` only if it is longer than the otherwise used value: Ok. Closing in favor of #12871

I've created a branch [jay:curl:crtdbg](https://github.com/curl/curl/compare/master...jay:curl:crtdbg?expand=1) that makes it possible to use Windows CRT heap memory tracking to detect leaks in libcurl *and* the application, with filenames and line numbers recorded...

Possibly related #5282. In short if you use OpenSSL as a static library then you may have to call [OPENSSL_thread_stop](https://www.openssl.org/docs/man3.0/man3/OPENSSL_thread_stop.html): _"Similarly this message will also not be sent if OpenSSL...

> Thanks, we'll test it today, will let you know later. Hope we can also get a repro for you or somehow debug why the issue still persists on shutdown...

> @jay PR updated, should be ready to merge. Perhaps, "fixes https://github.com/curl/curl/issues/13509" should be removed from the commit message Thanks. I left it with the fixed tag because it did...