git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: "Carlo Marcelo Arenas Belón" <carenas@gmail.com>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, Michael J Gruber <git@grubix.eu>
Subject: Re: [PATCH] http.c: clear the 'finished' member once we are done with it
Date: Fri, 6 May 2022 22:40:17 -0700	[thread overview]
Message-ID: <20220507054017.fnvb6xisr6s7m2l5@carlos-mbp.lan> (raw)
In-Reply-To: <xmqqczgqjr8y.fsf_-_@gitster.g>

On Fri, May 06, 2022 at 02:17:01PM -0700, Junio C Hamano wrote:
> diff --git a/http.c b/http.c
> index 229da4d148..85437b1980 100644
> --- a/http.c
> +++ b/http.c
> @@ -1367,6 +1367,9 @@ void run_active_slot(struct active_request_slot *slot)
>  			select(max_fd+1, &readfds, &writefds, &excfds, &select_timeout);
>  		}
>  	}
> +
> +	if (slot->finished == &finished)
> +		slot->finished = NULL;

I am not completely sure yet (since I looked at it long ago and got
sidetracked) but I think this might be optimized out (at least by gcc12)
since it is technically UB, which is why it never "fixed" the warning.

the "correct" way to implement this would be to make "finished" a thread
local static, which is finally one good reason to support C99, but the
syntax to do so with Windows broke my first attempt at doing so and now
I can even find the code I used then which required a per platform macro
and was better looking than the following

Carlo
--- >8 ----
Date: Wed, 20 Apr 2022 23:25:55 -0700
Subject: [PATCH] http: avoid using out of scope pointers

baa7b67d091 (HTTP slot reuse fixes, 2006-03-10) introduces a way
to notify a curl thread that its slot is finished by using a pointer
to a stack variable from run_active_slot(), but then gcc 12 was
released and started rightfully to complain about it (-Wdangling-pointer).

Use instead a thread storage static variable which is safe to use
between threads since C99 and doesn't go out of scope, while being
functionally equivalent to the original code, and also remove the
workaround from 9c539d1027d (config.mak.dev: alternative workaround
to gcc 12 warning in http.c, 2022-04-15)

Signed-off-by: Carlo Marcelo Arenas Belón <carenas@gmail.com>
---
 config.mak.dev | 1 -
 http.c         | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

diff --git a/config.mak.dev b/config.mak.dev
index c3104f400b..335efd4620 100644
--- a/config.mak.dev
+++ b/config.mak.dev
@@ -68,7 +68,6 @@ endif
 # https://bugzilla.redhat.com/show_bug.cgi?id=2075786
 ifneq ($(filter gcc12,$(COMPILER_FEATURES)),)
 DEVELOPER_CFLAGS += -Wno-error=stringop-overread
-DEVELOPER_CFLAGS += -Wno-error=dangling-pointer
 endif
 
 GIT_TEST_PERL_FATAL_WARNINGS = YesPlease
diff --git a/http.c b/http.c
index 229da4d148..cb9acfca19 100644
--- a/http.c
+++ b/http.c
@@ -1327,7 +1327,7 @@ void run_active_slot(struct active_request_slot *slot)
 	fd_set excfds;
 	int max_fd;
 	struct timeval select_timeout;
-	int finished = 0;
+	static __thread int finished;
 
 	slot->finished = &finished;
 	while (!finished) {
-- 
2.36.0.352.g0cd7feaf86f

  reply	other threads:[~2022-05-07  5:41 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-06 18:04 [PATCH 0/2] quell a few gcc warnings Michael J Gruber
2022-05-06 18:04 ` [PATCH 1/2] dir.c: avoid gcc warning Michael J Gruber
2022-05-06 20:21   ` Junio C Hamano
2022-05-09 15:58     ` Taylor Blau
2022-05-07  6:14   ` Carlo Marcelo Arenas Belón
2022-05-06 18:04 ` [PATCH 2/2] http.c: " Michael J Gruber
2022-05-06 20:22   ` Junio C Hamano
2022-05-06 21:17     ` [PATCH] http.c: clear the 'finished' member once we are done with it Junio C Hamano
2022-05-07  5:40       ` Carlo Marcelo Arenas Belón [this message]
2022-05-07 18:42         ` Junio C Hamano
2022-05-07 19:11           ` Carlo Arenas
2022-05-23 21:58       ` Johannes Schindelin
2022-05-23 22:58         ` Junio C Hamano
2022-05-23 23:36           ` Junio C Hamano
2022-05-23 23:41           ` Johannes Schindelin
2022-05-24  0:02             ` Junio C Hamano
2022-05-24  6:31               ` Daniel Stenberg
2022-05-24 10:57                 ` Johannes Schindelin
2022-05-24 17:45                 ` Junio C Hamano
2022-05-26 14:15                   ` Daniel Stenberg
2022-05-24 11:03               ` Johannes Schindelin
2022-05-24 17:15                 ` Junio C Hamano
2022-05-24 20:16                   ` Carlo Marcelo Arenas Belón
2022-05-24 20:45                     ` Ævar Arnfjörð Bjarmason
2022-05-24 22:34                       ` Junio C Hamano
2022-05-25  9:08                         ` Michael J Gruber
2022-05-25 13:27                         ` Ævar Arnfjörð Bjarmason
2022-05-24 22:16                     ` Junio C Hamano
2022-05-24 23:19                     ` Junio C Hamano
2022-05-25  2:02                       ` Carlo Arenas
2022-05-24 20:38                   ` Ævar Arnfjörð Bjarmason
2022-05-24 22:28                     ` Junio C Hamano
2022-05-25 10:07                       ` Johannes Schindelin
2022-05-25 16:40                         ` Junio C Hamano
2022-05-06 20:41   ` [PATCH 2/2] http.c: avoid gcc warning Carlo Marcelo Arenas Belón
2022-05-09 11:22 ` [PATCH] detect-compiler: make detection independent of locale Michael J Gruber
2022-05-09 15:52   ` Junio C Hamano
2022-05-09 15:59     ` rsbecker

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220507054017.fnvb6xisr6s7m2l5@carlos-mbp.lan \
    --to=carenas@gmail.com \
    --cc=git@grubix.eu \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).