From: Lars Ingebrigtsen <larsi@gnus.org>
Date: Sat, 1 Oct 2022 12:50:22 +0000 (+0200)
Subject: Fix url-retrieve-synchronously on very short timeouts
X-Git-Tag: emacs-29.0.90~1856^2~105
X-Git-Url: http://git.eshelyaron.com/gitweb/?a=commitdiff_plain;h=41234a21bf7d6713491ce60501bbf2d211fbac8e;p=emacs.git

Fix url-retrieve-synchronously on very short timeouts

* lisp/url/url.el (url-retrieve-synchronously): Don't kill the
process buffer unless we made one (bug#58218).  This makes HTTP
free connection marking work on premature timeouts.
---

diff --git a/lisp/url/url.el b/lisp/url/url.el
index d08ff04eda9..b4ece5faeb2 100644
--- a/lisp/url/url.el
+++ b/lisp/url/url.el
@@ -280,7 +280,9 @@ how long to wait for a response before giving up."
               ;; Querying over consumer internet in the US takes 100
               ;; ms, so split the difference.
               (accept-process-output nil 0.05)))
-        (unless (eq data-buffer proc-buffer)
+        ;; Kill the process buffer on redirects.
+        (when (and data-buffer
+                   (not (eq data-buffer proc-buffer)))
           (let (kill-buffer-query-functions)
             (kill-buffer proc-buffer)))))
     data-buffer))