From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from aws-us-west-2-korg-lkml-1.web.codeaurora.org (localhost.localdomain [127.0.0.1]) by smtp.lore.kernel.org (Postfix) with ESMTP id 41A80CA1010 for ; Wed, 3 Sep 2025 11:49:42 +0000 (UTC) Subject: Re: [walnascar 3/3] bitbake.conf: Switch prefix mapping to use -ffile-prefix-map To: openembedded-core@lists.openembedded.org From: "Varatharajan, Deepesh" X-Originating-Location: Bengaluru, Karnataka, IN (49.204.85.206) X-Originating-Platform: Windows Chrome 139 User-Agent: GROUPS.IO Web Poster MIME-Version: 1.0 Date: Wed, 03 Sep 2025 04:49:40 -0700 References: In-Reply-To: Message-ID: <4495.1756900180442086216@lists.openembedded.org> Content-Type: multipart/alternative; boundary="U7sT3ZBUYN3kJqWEn6Ew" List-Id: X-Webhook-Received: from li982-79.members.linode.com [45.33.32.79] by aws-us-west-2-korg-lkml-1.web.codeaurora.org with HTTPS for ; Wed, 03 Sep 2025 11:49:42 -0000 X-Groupsio-URL: https://lists.openembedded.org/g/openembedded-core/message/222856 --U7sT3ZBUYN3kJqWEn6Ew Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Hi Khem, We backported this patch due to a specific use case on our end. We set the = TMPDIR path length to 399 characters (just under the 400-character limit) and atte= mpted to build lib32-gcc. While this works fine on the master branch, we encounte= red the following error on the walnascar branch: make: /bin/bash: Argument list too long make: *** [Makefile:4701: all-gcc] Error 127 ERROR: oe_runmake failed Upon comparing the logs between the master and walnascar branches, we notic= ed that in walnascar, the following flags were passed multiple times: -fmacro-prefix-map -fdebug-prefix-map -fcanon-prefix-map Since our TMPDIR path is quite long, passing it repeatedly via these flags = seems to be the root cause of the issue. In the master branch, these were replaced with the more concise "-ffile-pre= fix-map" flag, significantly reducing the number of times the long TMPDIR path is pa= ssed to the compiler. This change resolved the error, which is why we backported th= e patch to walnascar. So, can we backport this for walnascar branch in future or do you suggest u= s to wait for october release. Regards, Deepesh --U7sT3ZBUYN3kJqWEn6Ew Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable
Hi Khem,

We backported this patch due to a specific use cas= e on our end. We set the TMPDIR
path length to 399 characters (just un= der the 400-character limit) and attempted
to build lib32-gcc. While t= his works fine on the master branch, we encountered the
following erro= r on the walnascar branch:

make: /bin/bash: Argument list too long
make: *** [Makefile:4701:= all-gcc] Error 127
ERROR: oe_runmake failed

Upon comparing the logs between the master and walnascar branches, we = noticed that
in walnascar, the following flags were passed multiple ti= mes:

-fmacro-prefix-map
-fdebug-prefix-map
-fcanon-prefix-map

Since our TMPDIR path is quite long, passing it repeatedly via these f= lags seems
to be the root cause of the issue.

In the master branch, these were replaced with the more concise "-ffil= e-prefix-map"
flag, significantly reducing the number of times the lon= g TMPDIR path is passed to
the compiler. This change resolved the erro= r, which is why we backported the patch
to walnascar. 

So, can we backport this for walnascar branch in future or do you suggest = us to wait for 
october release.

Regards,
Deepesh=

--U7sT3ZBUYN3kJqWEn6Ew--