3:08 a.m.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
6:46 a.m.
On Mon, 05 Nov 2001 12:08:40 +0900, "Stephen J. Turnbull"
<stephen(a)xemacs.org> said:
Lisp backtrace follows:
# bind (frame-being-created)
make-frame(nil #<x-device on ":0.0" 0x1bbe>)
frame-initialize()
# bind (debugger debug-on-error command-line-args-left)
command-line()
# (unwind-protect ...)
normal-top-level()
# (condition-case ... . error)
# (catch top-level ...)
(gdb) bt
#0 0x40227388 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
#1 0x402276bb in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
#2 0x40227876 in XtInitializeWidgetClass () from /usr/X11R6/lib/libXt.so.6
#3 0x40228154 in _XtCreatePopupShell () from /usr/X11R6/lib/libXt.so.6
#4 0x402281e5 in XtCreatePopupShell () from /usr/X11R6/lib/libXt.so.6
#5 0x081700f8 in x_create_widgets (f=3D0x8469480, lisp_window_id=3D1363249=
64, parent=3D136324964)
at frame-x.c:1957
#6 0x08170704 in x_init_frame_1 (f=3D0x8469480, props=3D136324964) at fram=
e-x.c:2184
#7 0x080f874d in Fmake_frame (props=3D136324964, device=3D138290840) at fr=
ame.c:395
Am seeing this as well, after updating to the latest (20011214) RedHat Rawhide.
I am *suspecting*, but cannot prove, that the binutils-2.11.92.0.12-3 RPM
is at fault, mostly because over on the Enlightenment list, 2.11.92 gave
some Debian users pain as well. 2.11.90 seems to be OK based on word from
that list.
Also, compiling with --pdump works for me as well - probably because this
way it avoids the unexec() that seems to be giving binutils indigestion.
One for the PROBLEMS file?
9:43 a.m.
On Sun, 16 Dec 2001 01:46:18 EST, Valdis.Kletnieks(a)vt.edu said:
Also, compiling with --pdump works for me as well - probably because
this
way it avoids the unexec() that seems to be giving binutils indigestion.
Blegga. Another one for the PROBLEMS file, self-inflicted this time.
If I didn't live on the ragged edge of RawHide, life would be easier. ;)
The RedHat gcc-3.1-0.10 RPM won't compile dumper.c with any level of -O,
it leaves a dangling entry resulting in a unknown symbol .L<n> for some
value of <n> because it botches the branch table for the switch statement
in pdump_get_indirect_count(). Already reported to RedHat as bug 57574.
Just in case anybody else is on the bleeding edge. ;)
/Valdis
3:41 a.m.
>>>> "Valdis" == Valdis Kletnieks
<Valdis.Kletnieks(a)vt.edu> writes:
> Lisp backtrace follows:
>
> # bind (frame-being-created)
This is the ld -z combreloc bug. Fixed in 21.4.6 and 21.5.4.
Use --pdump as a workaround.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
6:49 a.m.
On Mon, 17 Dec 2001 12:41:39 +0900, "Stephen J. Turnbull" said:
>>>>> "Valdis" == Valdis Kletnieks
<Valdis.Kletnieks(a)vt.edu> writes:
>> Lisp backtrace follows:
>>
>> # bind (frame-being-created)
This is the ld -z combreloc bug. Fixed in 21.4.6 and 21.5.4.
Use --pdump as a workaround.
I dug back a bit and found this from Nov 22 (the 21.4.6 notes):
4) A fix, actually two of them, for the "crash mysteriously on making
first frame" bug. GNU ld now literally stirs up your binary
before giving it back to you, leaving you fried. You can either
pass -z nocombreloc to ld, or use --pdump. Autoconfiscated,
should "just work". Bleeding edge Linux users, please test.
Is the autoconfiscation not in the 21.5 tree, or is it broken?
It certainly didnt "just work" with a CVS 21.5 from the last 48 hours
or so...
/Valdis
8:41 a.m.
>>>> "Valdis" == Valdis Kletnieks
<Valdis.Kletnieks(a)vt.edu> writes:
Valdis> Is the autoconfiscation not in the 21.5 tree, or is it
Valdis> broken? It certainly didnt "just work" with a CVS 21.5
Valdis> from the last 48 hours or so...
Not, I expect. I've spent little time on 21.5, and there are other
things (many) that need to be synched.
I plan to synch up 21.5 with the relevant changes to 21.4, and release
another 21.5 in the near future.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
4:42 p.m.
On Mon, 17 Dec 2001 17:41:04 +0900, "Stephen J. Turnbull" said:
>>>>> "Valdis" == Valdis Kletnieks
<Valdis.Kletnieks(a)vt.edu> writes:
Valdis> Is the autoconfiscation not in the 21.5 tree, or is it
Valdis> broken? It certainly didnt "just work" with a CVS 21.5
Valdis> from the last 48 hours or so...
Not, I expect. I've spent little time on 21.5, and there are other
things (many) that need to be synched.
I plan to synch up 21.5 with the relevant changes to 21.4, and release
another 21.5 in the near future.
OK. I'll keep an eye open on the CVS commits list, and test it when
it shows up.
/Valdis (who spent the weekend finding bugs in *everything*, it seems ;)
8159
days inactive
8201
days old
6 comments
2 participants
participants (2)
-
Stephen J. Turnbull -
Valdis Kl=?iso8859-4?Q?=BA?=tnieks