Bug 4248 - Host gcc 4.8 issues (gcc-cross-initial do_compile failure)
Summary: Host gcc 4.8 issues (gcc-cross-initial do_compile failure)
Status: RESOLVED FIXED
Alias: None
Product: OE-Core
Classification: Build System, Metadata & Runtime
Component: devtools / tool chain (show other bugs)
Version: unspecified
Hardware: x86 Multiple
: Medium+ normal
Target Milestone: 1.5
Assignee: Bogdan Marinescu
QA Contact:
URL:
Whiteboard: Release note for 1.4
Depends on:
Blocks:
 
Reported: 2013-04-11 11:41 UTC by Jack Mitchell
Modified: 2013-05-03 15:43 UTC (History)
4 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: ---

Attachments
do_compile log failure (901.69 KB, application/octet-stream)
2013-04-11 11:41 UTC, Jack Mitchell 		no flags Details
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Jack Mitchell 2013-04-11 11:41:22 UTC 
Created attachment 1162 [details]
do_compile log failure

GCC recently started to fail to build (3/4 days ago). Please see attached log. This is from a clean tmp and clean sstate.

Build Configuration:
BB_VERSION        = "1.19.0"
BUILD_SYS         = "x86_64-linux"
NATIVELSBSTRING   = "Unknown"
TARGET_SYS        = "arm-oecore-linux-gnueabi"
MACHINE           = "beaglebone"
DISTRO            = "dblinux"
DISTRO_VERSION    = "1.0-20130411"
TUNE_FEATURES     = "armv7a vfp neon"
TARGET_FPU        = "vfp-neon"
meta              = "master:a8188169eca8921320852353a4afa2e99c0a2336"
meta-webserver    = "master:222b4eb3e469a84e929a968def6773405f369a3d"
meta-db           = "master:a7f21669983c1e9e3cac479f0d1ee017eb50f992"
meta-dblinux      = "master:36e8caf34b5f5e07d92d544dd523b33fabe184b4"
common-bsp        = "master:61d0b1690637966a9716ea067c050beeac97943e"
Comment 1 Jack Mitchell 2013-04-11 12:56:39 UTC 
With Khem's GCC 4.8 patches I have no issues. I am happy to use them and use GCC 4.8 but I doubt that they will make it into 1.4 so I suppose this is still an issue...
Comment 2 Richard Purdie 2013-04-11 15:59:55 UTC 
Are you able to bisect when this started happening? There haven't been any changes in gcc for a while, certainly not the last 3/4 days so this is a bit puzzling.
Comment 3 Paul Eggleton 2013-04-11 16:31:03 UTC 
This is the host gcc ICEing, which we can't do a lot about. Did you get upgraded to gcc 4.8.0 on your host by any chance? We've had reports on IRC of  this exact issue with 4.8.0 on the host (with Arch Linux in at least once case).
Comment 4 Jack Mitchell 2013-04-16 08:30:12 UTC 
Sorry for the late reply. Yes, this is on ArchLinux which does indeed seem to have a fault; I have moved to Khems 4.8 patches now, so it will be a while before I can test this again, however it seems other people are now popping up with the issue, so maybe they can provide more input if they require gcc 4.7.

ArchLinux provided a new package release of GCC yesterday which may contain a fix.
Comment 5 Paul Eggleton 2013-05-03 15:43:47 UTC 
This should now be fixed in master (and will be backported to dylan):

http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=3681c89432b5fb705fe593e75c2d98f7d15a5184