public inbox for dwz@sourceware.org
 help / color / mirror / Atom feed
From: Tom de Vries <tdevries@suse.de>
To: Mark Wielaard <mark@klomp.org>, dwz@sourceware.org
Subject: Re: Buildbot failure in Wildebeest Builder on whole buildset
Date: Tue, 01 Jan 2019 00:00:00 -0000	[thread overview]
Message-ID: <6db21006-1eed-1224-bc42-a665a757dc05@suse.de> (raw)
In-Reply-To: <8f249c07c1cb9a6cbf78eec5c05b3c5641cd5526.camel@klomp.org>

[-- Attachment #1: Type: text/plain, Size: 1591 bytes --]

On 30-03-19 00:42, Mark Wielaard wrote:
> On Fri, 2019-03-29 at 21:42 +0000, buildbot@builder.wildebeest.org
> wrote:
>> The Buildbot has detected a failed build on builder whole buildset
>> while building dwz.
>> Full details are available at:
>>     https://builder.wildebeest.org/buildbot/#builders/19/builds/34
>>
>> Buildbot URL: https://builder.wildebeest.org/buildbot/
>>
>> Worker for this Build: fedora-x86_64
> 
> Which is Fedora 29. The other builders that fail are also Fedora based.
> The Debian and CentOS builds still seem to work.
> 
> The problem is this commit:
> 
> commit 388977c1ccdaf376540a651ee34cdf709891e0fe
> Author: Tom de Vries <tdevries@suse.de>
> Date:   Fri Mar 29 21:55:24 2019 +0100
> 
>     Compile dwz-for-tests with -U__GNUC__
> 
> Undefining __GNUC__ causes issues with some include files,
> specifically:
> 
> In file included from /usr/include/bits/floatn.h:119,
>                  from /usr/include/stdlib.h:55,
>                  from dwz.c:30
> 
> __GNUC__ defines the major version of GCC (8 on Fedora 29).
> 
> Since GCC 7 the compiler defines the builtin types _Float32 and
> _Float64 so floatn.h doesn't have to define them. But since we undefine
> __GNUC__ the header cannot check the GCC version anymore and tries to
> define them anyway. Causing the compile error.
> 

Thanks for the investigation.

> I think the commit should just be reverted.
> It isn't clear to me what it really tries to check for.

I'm trying to make sure that dwz builds in !__GNUC__ mode.

Fixed in attached commit, buildbot status back to all clear.

Thanks,
- Tom


[-- Attachment #2: 0001-Fix-dwz-for-test-build-failure-on-fedora.patch --]
[-- Type: text/x-patch, Size: 1699 bytes --]

Fix dwz-for-test build failure on fedora

Commit "Compile dwz-for-tests with -U__GNUC__" makes sure we we trigger the
!__GNUC__ path.  But -U__GNUC__ makes compilation of dwz-for-test failure on
fedora.

Instead, generate sources for dwz-for-test with __GNUC__ replaced by
NOT_DEFINED.

2019-03-30  Tom de Vries  <tdevries@suse.de>

	* Makefile (DWZ_TEST_SOURCES): New variable.
	(%-for-test.c): New target.  Generate %-for-test.c from %.c with
	__GNUC__ replaced by NOT_DEFINED.
	(dwz-for-test): Remove -U__GNUC__.  Use DWZ_TEST_SOURCES.
	(check): Remove DWZ_TEST_SOURCES when done.

---
 Makefile | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/Makefile b/Makefile
index 0ce4482..7150d4a 100644
--- a/Makefile
+++ b/Makefile
@@ -29,9 +29,14 @@ dw2-restrict:
 py-section-script:
 	$(CC) $(TEST_SRC)/py-section-script.s -o $@ -g || touch $@
 
-dwz-for-test:
-	$(CC) $(patsubst %.o,%.c,$(OBJECTS)) -O2 -g -lelf -o $@ -Wall -W \
-	  -D_FILE_OFFSET_BITS=64 -DDWZ_VERSION='"for-test"' -U__GNUC__
+DWZ_TEST_SOURCES := $(patsubst %.o,%-for-test.c,$(OBJECTS))
+
+%-for-test.c: %.c
+	sed 's/__GNUC__/NOT_DEFINED/' $< > $@
+
+dwz-for-test: $(DWZ_TEST_SOURCES)
+	$(CC) $(DWZ_TEST_SOURCES) -O2 -g -lelf -o $@ -Wall -W \
+	  -D_FILE_OFFSET_BITS=64 -DDWZ_VERSION='"for-test"'
 
 # On some systems we need to set and export DEJAGNU to suppress
 # WARNING: Couldn't find the global config file.
@@ -43,4 +48,4 @@ check: dwz $(TEST_EXECS)
 	export DEJAGNU=$(DEJAGNU); \
 	export PATH=$(PWD)/testsuite-bin:$$PATH; export LC_ALL=C; \
 	runtest --tool=dwz -srcdir testsuite $(RUNTESTFLAGS)
-	rm -Rf testsuite-bin $(TEST_EXECS)
+	rm -Rf testsuite-bin $(TEST_EXECS) $(DWZ_TEST_SOURCES)

  reply	other threads:[~2019-03-30  7:42 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-01  0:00 buildbot
2019-01-01  0:00 ` Mark Wielaard
2019-01-01  0:00   ` Tom de Vries [this message]
  -- strict thread matches above, loose matches on Subject: below --
2021-03-26 13:04 buildbot
2021-03-25  8:41 buildbot
2020-10-13 21:29 buildbot
2020-10-13 22:20 ` Mark Wielaard
2020-10-14 10:36   ` Mark Wielaard
2020-07-19 23:00 buildbot
2020-07-19 23:17 ` Mark Wielaard
2020-01-01  0:00 buildbot
2020-01-01  0:00 ` Mark Wielaard
2019-01-01  0:00 buildbot
2019-01-01  0:00 ` Mark Wielaard
2019-01-01  0:00 buildbot
2019-01-01  0:00 buildbot
2019-01-01  0:00 buildbot
2019-01-01  0:00 buildbot
2019-01-01  0:00 buildbot
2019-01-01  0:00 buildbot

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

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

  git send-email \
    --in-reply-to=6db21006-1eed-1224-bc42-a665a757dc05@suse.de \
    --to=tdevries@suse.de \
    --cc=dwz@sourceware.org \
    --cc=mark@klomp.org \
    /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.
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).