Summary: | PowerPC64 changes break makedepend/imake on non-PPC64 platforms | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Alan Coopersmith <alan.coopersmith> | ||||||
Component: | Build/Monolithic | Assignee: | Mike A. Harris <mharris> | ||||||
Status: | RESOLVED FIXED | QA Contact: | |||||||
Severity: | normal | ||||||||
Priority: | high | CC: | eich, kem, mharris | ||||||
Version: | unspecified | ||||||||
Hardware: | x86 (IA32) | ||||||||
OS: | All | ||||||||
Whiteboard: | |||||||||
i915 platform: | i915 features: | ||||||||
Bug Depends on: | |||||||||
Bug Blocks: | 351 | ||||||||
Attachments: |
|
Description
Alan Coopersmith
2004-03-12 17:08:29 UTC
Indeed, something looks awry there. Just as a datapoint what OS/version and compiler/version are you using? The error you describe doesn't occur here, but from the look of the code it should. I'll look more closely into this on Monday. Thanks for spotting this Alan. I saw this on Solaris 9 x86 with the Sun compilers. This patch shouldn't be in the RELEASE branch anyway. Please fix and apply it to the CURRENT branch. Assign it to the next release. Created attachment 175 [details] [review] Updated PPC64 support patch This updated patch includes the support which was previously committed by me to the XORG-RELEASE-1 branch and later reverted, plus fixes for the issues reported by Alan Coopersmith in this bug report, and an additional fix for an issue we discovered in routine testing, where BIG_ENDIAN wasn't getting properly defined on PPC64 with the X.org release. The patch is also updated to apply cleanly to the current CVS as of a couple of minutes ago, and should apply without fuzz, etc. I'm posting it here first for review by Alan and Egbert, and would prefer to get at least 2 people to sign off on it before I commit it to XORG-CURRENT in CVS. Any build/runtime testing others would be willing to do would also be appreciated, in particular on non-Linux systems, and with non GNU compilers, as that improves the chances of problems getting detected. If there are any further issues with the patch, please let me know and I will try to address them and provide an updated patch once it can be tested. Thanks in advance. Created attachment 183 [details] [review] PPC64 support patch, updated to work with final X11R6.7.0 release This patch, has been updated to work with final X11R6.7.0 release. It is identical, except changed to patch xorg.cf instead of xfree86.cf. The latest patch looks fine to me, and I don't see any of the previous errors when I apply to a checkout from the current tree and build it on the Solaris x86 system that previously hit errors. I don't have a PPC or PPC64 box to confirm it does the right thing there - but since it seems to not break existing platforms, I think it should be safe to check in now. Shall I check it into the trunk now (HEAD), or is some other branch more appropriate? Just want to confirm first. Has anyone tested this patch against the current trunk? Does anyone have a PPC64 system to test? The patch looks innocuous enough but it should be tested before checking into the trunk. I'm currently build testing ppc64 in FC3 devel builds. s390 is failing, so not sure if ppc64 builds clean yet or not. Will know over the weekend. This patch is applied in our rpms, and should be harmless to apply IMHO unless someone knows of issues on non-Linux systems, but nothing has been reported to date that I'm aware. Once I get a successful xorg ppc64 build, I'll update the status here. Any updates on the build status with the current CVS head? The current 6.7.99.x src.rpm builds on PPC64 just fine now. Any of the build problems I encountered while packaging this for PPC64 were minor rpm packaging related issues. There are no PPC64 or other arch/OS related issues that I'm aware of currently with this patch applied to the Xorg sources. I believe it should be safe to apply. Patch checked in. Closing. Changed xorg.cf to define XorgServer instead of XF86Server. From mharris. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.