Skip to contentSkip to conversation

binutils-native 2.24 build broken

DISCUSSION Darcy Watkins 2 participants 6 messages Jul 28, 2014

lore
Share & tools
lore
#0 lore raw

Hi,

I encounter the following while building based on 'daisy':

| gcc  -DHAVE_CONFIG_H -I. -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld  -I. -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld -I../bfd -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld/../bfd -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld/../include  -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -O2 -pipe -DENABLE_PLUGINS -DLOCALEDIR="\"/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/share/locale\"" -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Wshadow -Werror -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -O2 -pipe -c eavr25.c
| gcc  -DHAVE_CONFIG_H -I. -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld  -I. -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld -I../bfd -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld/../bfd -I/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/ld/../include  -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -O2 -pipe -DENABLE_PLUGINS -DLOCALEDIR="\"/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/share/locale\"" -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Wshadow -Werror -isystem/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/sysroots/i686-linux/usr/include -O2 -pipe -c eavr3.c
| cc1: warnings being treated as errors
| eavr2.c: In function 'avr_elf_avr2_before_allocation':
| eavr2.c:1200: error: 'ehdr_start_save$type' may be used uninitialized in this function
| eavr2.c:1200: note: 'ehdr_start_save$type' was declared here
| make[4]: *** [eavr2.o] Error 1
| make[4]: *** Waiting for unfinished jobs....
| cc1: warnings being treated as errors
| eavr1.c: In function 'avr_elf_avr1_before_allocation':
| eavr1.c:1200: error: 'ehdr_start_save$type' may be used uninitialized in this function
| eavr1.c:1200: note: 'ehdr_start_save$type' was declared here
| make[4]: *** [eavr1.o] Error 1
| cc1: warnings being treated as errors
| eavr25.c: In function 'avr_elf_avr25_before_allocation':
| eavr25.c:1200: error: 'ehdr_start_save$type' may be used uninitialized in this function
| eavr25.c:1200: note: 'ehdr_start_save$type' was declared here
| make[4]: *** [eavr25.o] Error 1
| cc1: warnings being treated as errors
| eavr3.c: In function 'avr_elf_avr3_before_allocation':
| eavr3.c:1200: error: 'ehdr_start_save$type' may be used uninitialized in this function
| eavr3.c:1200: note: 'ehdr_start_save$type' was declared here
| make[4]: *** [eavr3.o] Error 1
| make[4]: Leaving directory `/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/build.i686-linux.i686-linux/ld'
| make[3]: *** [all-recursive] Error 1
| make[3]: Leaving directory `/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/build.i686-linux.i686-linux/ld'
| make[2]: *** [all] Error 2
| make[2]: Leaving directory `/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/build.i686-linux.i686-linux/ld'
| make[1]: *** [all-ld] Error 2
| make[1]: Leaving directory `/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/binutils-2.24/build.i686-linux.i686-linux'
| make: *** [all] Error 2
| ERROR: oe_runmake failed
| WARNING: /home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/temp/run.do_compile.11497:1 exit 1 from
|   exit 1
| ERROR: Function failed: do_compile (log file is located at /home/dwatkins/workspace/epsilon/daisy_MG-os_test2/build/tmp/work/i686-linux/binutils-native/2.24-r0/temp/log.do_compile.11497)

ERROR: Task 780 (virtual:native:/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/sources/poky/meta/recipes-devtools/binutils/binutils_2.24.bb, do_compile) failed with exit code '1'
NOTE: Tasks Summary: Attempted 2208 tasks of which 2150 didn't need to be rerun and 1 failed.
Waiting for 0 running tasks to finish:

Summary: 1 task failed:
virtual:native:/home/dwatkins/workspace/epsilon/daisy_MG-os_test2/sources/poky/meta/recipes-devtools/binutils/binutils_2.24.bb, do_compile
Summary: There were 9 WARNING messages shown.
Summary: There was 1 ERROR message shown, returning a non-zero exit code.

I use a layout much like Freescale Community BSP project that uses 'repo' to pull together the layers based on a manifest.

I noticed this when a syncup triggered the toolchain to rebuild.

Since I hadn't had anything trigger toolchain to rebuild since July 17, I tried 'git checkout' all my layers, oe layers and yocto layers to specific hashes to get back to the state I had in July 17. I blew away the sstate-cache, cache and tmp and rebuilt from scratch. Same thing.

Are there floating git references related to the toolchain that could have updated independent of the meta data that could still go forward of Jul 17?

In any case, it appears that my 64bit Fedora machine at home is not affected by this, but my 32bit CentOS 6 machine at work is affected by it.

Anyone know what happened?

Signature · 13 lines
Regards,

Darcy

---

Darcy Watkins
Staff Engineer, Firmware
In Motion Technology, a
Sierra Wireless company
350-625 Agnes St. New Westminster, BC
Canada, V3M 5Y4
[P1]
#1 lore raw
Quoted 15 lines · Darcy Watkins

I use a layout much like Freescale Community BSP project that uses 'repo' to pull together the layers based on a manifest.

I noticed this when a syncup triggered the toolchain to rebuild.

Since I hadn't had anything trigger toolchain to rebuild since July 17, I tried 'git checkout' all my layers, oe layers and yocto layers to specific hashes to get back to the state I had in July 17. I blew away the sstate-cache, cache and tmp and rebuilt from scratch. Same thing.

Are there floating git references related to the toolchain that could have updated independent of the meta data that could still go forward of Jul 17?

Quoted 5 lines

In any case, it appears that my 64bit Fedora machine at home is not affected by this, but my 32bit CentOS 6 machine at work is affected by it.

Anyone know what happened?

can you apply the following patch to binutils and retry

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=f9c316c4d75be236bbaa8464ef803ed2d3859d6d

#2 lore raw
Quoted 3 lines · Khem Raj

can you apply the following patch to binutils and retry

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=f9c316c4d75be236bbaa8464ef803ed2d3859d6d

The patch appears to have worked to get me past binutils-native into the rest of the build. I have placed it with a bbappend file in a hotfix section of my layer.

Please let me know once this fix makes it way into the poky/daisy branch.

Thanks!

Signature · 13 lines
Regards,

Darcy

---

Darcy Watkins
Staff Engineer, Firmware
In Motion Technology, a
Sierra Wireless company
350-625 Agnes St. New Westminster, BC
Canada, V3M 5Y4
[P1]
#3 lore raw
Quoted 6 lines · Darcy Watkins

The patch appears to have worked to get me past binutils-native into the rest of the build. I have placed it with a bbappend file in a hotfix section of my layer.

Please let me know once this fix makes it way into the poky/daisy branch.

If you can please submit the patch to OE-Core mailing list as well for master and also a backport to daisy please.

#4 lore raw
Quoted 2 lines · Khem Raj

If you can please submit the patch to OE-Core mailing list as well for master and also a backport to daisy please.

I found the patch is already present in poky/master so it is just the backport to daisy that is needed (possibly other branches as well).

commit 540b7a196a00c4dc1bf8f7c87e68e80ab310756e

Author: Richard Purdie <[email protected]>
Date:   Wed Jun 25 09:09:44 2014 +0000

binutils: Add fix for recent patch on older gcc

The recent bintuils patch update breaks on older gccs such as CentOS 6. Backport a patch to address this.

(From OE-Core rev: 76f65e73081f52cea718ef164f9d1d7a5c65d537)

This commit applies cleanly as-is to the daisy branch. Do I need to resubmit the patch by Richard Purdie, or does maintainer prefer to merge the commit across branches?

Signature · 13 lines
Regards,

Darcy

---

Darcy Watkins
Staff Engineer, Firmware
In Motion Technology, a
Sierra Wireless company
350-625 Agnes St. New Westminster, BC
Canada, V3M 5Y4
[P1]
#5 lore raw
Quoted 3 lines · Darcy Watkins

This commit applies cleanly as-is to the daisy branch. Do I need to resubmit the patch by Richard Purdie, or does maintainer prefer to merge the commit across branches?

Backport Richard's please