Bug 744740 - pth fails to build on Linux 3.x
Summary: pth fails to build on Linux 3.x
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: pth
Version: 19
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Michael Schwendt
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-10-10 11:04 UTC by Daniel Drake
Modified: 2013-09-22 16:47 UTC (History)
1 user (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2013-09-22 16:47:34 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
fix pth for makecontext-less glibc on Linux 3.x kernels (4.28 KB, patch)
2012-06-14 12:35 UTC, Mikael Pettersson
no flags Details | Diff

Description Daniel Drake 2011-10-10 11:04:50 UTC
I'd file this directly upstream, but mails to the mailing list subscription and bug addresses are bouncing.

Build fails on armv5tel at least because:

./libtool --mode=compile --quiet gcc -c -I. -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -march=armv5te -mfloat-abi=soft pth_mctx.c
pth_mctx.c: In function '__pth_mctx_set':
pth_mctx.c:480:2: error: #error "Unsupported Linux (g)libc version and/or platform"
make: *** [pth_mctx.lo] Error 1


Let me save you a headache: the (obscure) problem here is that it doesn't like Linux 3.x as a build host. 
aclocal.m4 does:

case $PLATFORM in
    *-*-linux* )
        braindead=no
        case "x`uname -r`" in
changequote(, )dnl
            x2.[23456789]* ) ;;
changequote([, ])
            * ) braindead=yes ;;


it therefore thinks Linux 3.x is braindead, which is the first issue that leads to the above compile error.

Comment 1 Michael Schwendt 2011-10-10 12:07:19 UTC
Something about your observation can't be true, because recently, pth has been rebuilt in Rawhide with a Linux kernel 3.1.0-0.rc6.git0.0.fc17:
http://koji.fedoraproject.org/koji/buildinfo?buildID=263955

As I don't have access to any ARMv5TE machine, I cannot examine this myself.

Comment 2 Daniel Drake 2011-10-10 12:12:02 UTC
I don't think that was built on a Linux 3.1 host. root.log says:
DEBUG backend.py:394:  kver == 2.6.32-131.2.1.el6.x86_64

Comment 3 Michael Schwendt 2011-10-10 12:18:16 UTC
And still pth builds fine in F-16 x86_64 here, too:
$ uname -r
3.1.0-0.rc9.git0.0.fc16.x86_64

Comment 4 Daniel Drake 2011-10-10 12:23:51 UTC
Ah yes. Looking closer, this bug will only affect architectures that do not have the MCSC method. x86/x86_64 has that method, ARM falls back on SJLJ, and it is in the SJLJ code where braindeath occurs :)

Comment 5 Mikael Pettersson 2012-06-14 12:35:33 UTC
Created attachment 591825 [details]
fix pth for makecontext-less glibc on Linux 3.x kernels

I faced the same problem (pth build breaks with a Linux 3.x kernel and a glibc that lacks makecontext et al) recently in my private distro.  I've been trying to submit this fix to pth upstream but my emails seem to end up in /dev/null.  If Fedora could carry the fix for the time being that would be great.

Comment 6 Michael Schwendt 2012-06-16 16:08:02 UTC
Applied in Rawhide.

Comment 7 Fedora End Of Life 2013-04-03 16:25:27 UTC
This bug appears to have been reported against 'rawhide' during the Fedora 19 development cycle.
Changing version to '19'.

(As we did not run this process for some time, it could affect also pre-Fedora 19 development
cycle bugs. We are very sorry. It will help us with cleanup during Fedora 19 End Of Life. Thank you.)

More information and reason for this action is here:
https://fedoraproject.org/wiki/BugZappers/HouseKeeping/Fedora19


Note You need to log in before you can comment on or make changes to this bug.