FTBFS (alpha): autogen segfaults on license.test
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
autogen (Debian) |
Fix Released
|
Unknown
|
|||
autogen (Ubuntu) |
Invalid
|
High
|
Unassigned |
Bug Description
Automatically imported from Debian bug report #340851 http://
Debian Bug Importer (debzilla) wrote : | #1 |
Debian Bug Importer (debzilla) wrote : | #2 |
Message-Id: <E1EfyWo-
Date: Sat, 26 Nov 2005 12:47:38 +0100
From: Falk Hueffner <email address hidden>
To: Debian Bug Tracking System <email address hidden>
Subject: FTBFS (alpha): autogen segfaults on license.test
Package: autogen
Version: 1:5.7.2-1
Severity: serious
Justification: no longer builds from source
autogen does not build on Alpha:
[...]
FAIL: license.test
[...]
=======
1 of 34 tests failed
Please report to <email address hidden>
=======
Full log at
http://
The reason is that autogen segfaults on this example.
juist:/
Segmentation fault (core dumped)
The backtrace is not very helpful, unfortunately:
juist:/
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB. Type "show warranty" for details.
This GDB was configured as "alpha-
(gdb) r -b license --no-def -T license.tpl
Starting program: /tmp/autogen-
[Thread debugging using libthread_db enabled]
[New Thread 16384 (LWP 27322)]
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 27322)]
0x0000020000008a68 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
(gdb) bt
#0 0x0000020000008a68 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#1 0x0000020000008efc in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#2 0x000002000000c540 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#3 0x000002000000c058 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#4 0x000000012001a9fc in ag_scm_license ()
#5 0x00000001200930b0 in ?? ()
warning: Hit heuristic-
This warning occurs if you are debugging a function without any symbols
(for example, in a stripped executable). In that case, you may wish to
increase the size of the search with the `set heuristic-
Otherwise, you told GDB there was a function where there isn't one, or
(more likely) you have encountered a bug in GDB.
#6 0x00000001200930b0 in ?? ()
warning: Hit heuristic-
Previous frame identical to this frame (corrupt stack?)
(gdb)
It also fails with autogen 1:5.7.2-1 from sid:
root@juist:
AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
processing template NULL file name
on line 0
for function pseudo-macro (-1)
zsh: abort (core dumped) autogen -b license --no-def -...
Dennis Kaarsemaker (dennis) wrote : | #3 |
-EARCH
In Debian Bug tracker #340851, Bruce Korb (bruce-korb-symantec) wrote : 5.7.3 fixed two license test bugs | #4 |
1. the test used a POSIX-ism that failed if the shell did not
"grok" POSIX extensions to Bourne.
2. If the license file size was a multiple of pagesize, then
strlen() would seg fault.
OTOH, neither of these would affect the pseudo macro processing:
> It also fails with autogen 1:5.7.2-1 from sid:
>
> root@juist:
> autogen -b license --no-def -T license.tpl
> AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> processing template NULL file name
> on line 0
> for function pseudo-macro (-1)
> zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
It does appear though that once you have a failing example, it fails
regularly. (You caused this by typing in the command, yes?). Maybe
I can debug the issue with the contents of the FAILURES directory?
Any help would be appreciated. Thanks - Bruce
P.S. I'm curious:
> Package: autogen
> Version: 1:5.7.2-1
> Severity: serious
> Justification: no longer builds from source
yet I just noticed the "autogen-5.7.3" element in the path above.
If you were using 5.7.3 and encountered this problem, then the
fixes noted above are already in the sources you were working with....
Debian Bug Importer (debzilla) wrote : | #5 |
Message-Id: <email address hidden>
Date: Sat, 26 Nov 2005 09:12:31 -0800
From: Bruce Korb <email address hidden>
To: <email address hidden>
Subject: 5.7.3 fixed two license test bugs
1. the test used a POSIX-ism that failed if the shell did not
"grok" POSIX extensions to Bourne.
2. If the license file size was a multiple of pagesize, then
strlen() would seg fault.
OTOH, neither of these would affect the pseudo macro processing:
> It also fails with autogen 1:5.7.2-1 from sid:
>
> root@juist:
> autogen -b license --no-def -T license.tpl
> AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> processing template NULL file name
> on line 0
> for function pseudo-macro (-1)
> zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
It does appear though that once you have a failing example, it fails
regularly. (You caused this by typing in the command, yes?). Maybe
I can debug the issue with the contents of the FAILURES directory?
Any help would be appreciated. Thanks - Bruce
P.S. I'm curious:
> Package: autogen
> Version: 1:5.7.2-1
> Severity: serious
> Justification: no longer builds from source
yet I just noticed the "autogen-5.7.3" element in the path above.
If you were using 5.7.3 and encountered this problem, then the
fixes noted above are already in the sources you were working with....
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : Re: Bug#340851: 5.7.3 fixed two license test bugs | #6 |
On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> P.S. I'm curious:
>
> > Package: autogen
> > Version: 1:5.7.2-1
> > Severity: serious
> > Justification: no longer builds from source
>
> yet I just noticed the "autogen-5.7.3" element in the path above.
> If you were using 5.7.3 and encountered this problem, then the
> fixes noted above are already in the sources you were working with....
According to the full build log, it was building 5.7.3-1.
--
Matt
Debian Bug Importer (debzilla) wrote : | #7 |
Message-ID: <email address hidden>
Date: Sat, 26 Nov 2005 14:39:01 -0800
From: Matt Kraai <email address hidden>
To: <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 fixed two license test bugs
--k4f25fnPtRuIRUb3
Content-Type: text/plain; charset=us-ascii
Content-
Content-
On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> P.S. I'm curious:
>=20
> > Package: autogen
> > Version: 1:5.7.2-1
> > Severity: serious
> > Justification: no longer builds from source
>=20
> yet I just noticed the "autogen-5.7.3" element in the path above.
> If you were using 5.7.3 and encountered this problem, then the
> fixes noted above are already in the sources you were working with....
According to the full build log, it was building 5.7.3-1.
--=20
Matt
--k4f25fnPtRuIRUb3
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDiOQEfNd
dkyTw2J438WAeeQ
=KA5o
-----END PGP SIGNATURE-----
--k4f25fnPtRuIR
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : | #8 |
On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> > It also fails with autogen 1:5.7.2-1 from sid:
> >
> > root@juist:
> > autogen -b license --no-def -T license.tpl
> > AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> > processing template NULL file name
> > on line 0
> > for function pseudo-macro (-1)
> > zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
>
> It does appear though that once you have a failing example, it fails
> regularly. (You caused this by typing in the command, yes?). Maybe
> I can debug the issue with the contents of the FAILURES directory?
I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
unstable chroot on escher.debian.org. Falk, would you try to
reproduce the problem and let me know whether or not you are able to?
--
Matt
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : tag 340851 unreproducible | #9 |
tag 340851 unreproducible
--
Matt
Debian Bug Importer (debzilla) wrote : | #10 |
Message-ID: <email address hidden>
Date: Mon, 28 Nov 2005 17:52:27 -0800
From: Matt Kraai <email address hidden>
To: <email address hidden>
Subject: tag 340851 unreproducible
--4SFOXa2GPu3tIq4H
Content-Type: text/plain; charset=us-ascii
Content-
Content-
tag 340851 unreproducible
--=20
Matt
--4SFOXa2GPu3tIq4H
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDi7RafNd
b8ug4c/
=CU9D
-----END PGP SIGNATURE-----
--4SFOXa2GPu3tI
Debian Bug Importer (debzilla) wrote : | #11 |
Message-ID: <email address hidden>
Date: Mon, 28 Nov 2005 17:51:43 -0800
From: Matt Kraai <email address hidden>
To: <email address hidden>
Cc: Bruce Korb <email address hidden>
Subject: Re: Bug#340851: 5.7.3 fixed two license test bugs
--jRHKVT23PllUwdXP
Content-Type: text/plain; charset=us-ascii
Content-
Content-
On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> > It also fails with autogen 1:5.7.2-1 from sid:
> >=20
> > root@juist:
FAILURES# \
> > autogen -b license --no-def -T license.tpl=20
> > AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> > processing template NULL file name
> > on line 0
> > for function pseudo-macro (-1)
> > zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
>=20
> It does appear though that once you have a failing example, it fails
> regularly. (You caused this by typing in the command, yes?). Maybe
> I can debug the issue with the contents of the FAILURES directory?
I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
unstable chroot on escher.debian.org. Falk, would you try to
reproduce the problem and let me know whether or not you are able to?
--=20
Matt
--jRHKVT23PllUwdXP
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDi7QufNd
2uSseIHhtnyihph
=IvBW
-----END PGP SIGNATURE-----
--jRHKVT23PllUw
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : Re: Bug#340851: 5.7.3 fixed two license test bugs | #12 |
On Tue, Nov 29, 2005 at 09:07:39AM +0100, Falk Hueffner wrote:
> Matt Kraai <email address hidden> writes:
>
> > On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> >> > It also fails with autogen 1:5.7.2-1 from sid:
> >> >
> >> > root@juist:
> >> > autogen -b license --no-def -T license.tpl
> >> > AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> >> > processing template NULL file name
> >> > on line 0
> >> > for function pseudo-macro (-1)
> >> > zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
> >>
> >> It does appear though that once you have a failing example, it fails
> >> regularly. (You caused this by typing in the command, yes?). Maybe
> >> I can debug the issue with the contents of the FAILURES directory?
> >
> > I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
> > unstable chroot on escher.debian.org. Falk, would you try to
> > reproduce the problem and let me know whether or not you are able to?
>
> I was able to reproduce it within pbuilder's chroot, but not outside
> (both sid). Weird. If anybody wants to debug this, I can try to set up
> an environment where it can be reproduced and provide an account.
If you can give me access to an environment where I can reproduce it,
I'll investigate further.
--
Matt
Debian Bug Importer (debzilla) wrote : | #13 |
Message-ID: <email address hidden>
Date: Tue, 29 Nov 2005 19:33:04 -0800
From: Matt Kraai <email address hidden>
To: Falk Hueffner <email address hidden>
Cc: <email address hidden>, Bruce Korb <email address hidden>
Subject: Re: Bug#340851: 5.7.3 fixed two license test bugs
--4bRzO86E/ozDv8r1
Content-Type: text/plain; charset=us-ascii
Content-
Content-
On Tue, Nov 29, 2005 at 09:07:39AM +0100, Falk Hueffner wrote:
> Matt Kraai <email address hidden> writes:
>=20
> > On Sat, Nov 26, 2005 at 09:12:31AM -0800, Bruce Korb wrote:
> >> > It also fails with autogen 1:5.7.2-1 from sid:
> >> >=20
> >> > root@juist:
st/FAILURES# \
> >> > autogen -b license --no-def -T license.tpl=20
> >> > AutoGen aborting on signal 10 (Bus error) in state LOAD_TPL
> >> > processing template NULL file name
> >> > on line 0
> >> > for function pseudo-macro (-1)
> >> > zsh: abort (core dumped) autogen -b license --no-def -T license.tpl
> >>=20
> >> It does appear though that once you have a failing example, it fails
> >> regularly. (You caused this by typing in the command, yes?). Maybe
> >> I can debug the issue with the contents of the FAILURES directory?
> >
> > I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
> > unstable chroot on escher.debian.org. Falk, would you try to
> > reproduce the problem and let me know whether or not you are able to?
>=20
> I was able to reproduce it within pbuilder's chroot, but not outside
> (both sid). Weird. If anybody wants to debug this, I can try to set up
> an environment where it can be reproduced and provide an account.
If you can give me access to an environment where I can reproduce it,
I'll investigate further.
--=20
Matt
--4bRzO86E/ozDv8r1
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDjR1vfNd
KMEZ+j7lRaleSnu
=O8PD
-----END PGP SIGNATURE-----
--4bRzO86E/
In Debian Bug tracker #340851, Bkorb (bkorb) wrote : | #14 |
On Tuesday 29 November 2005 07:33 pm, Matt Kraai wrote:
> > > I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
> > > unstable chroot on escher.debian.org. Falk, would you try to
> > > reproduce the problem and let me know whether or not you are able to?
> >
> > I was able to reproduce it within pbuilder's chroot, but not outside
> > (both sid). Weird. If anybody wants to debug this, I can try to set up
> > an environment where it can be reproduced and provide an account.
>
> If you can give me access to an environment where I can reproduce it,
> I'll investigate further.
Me, too!! :) It would be nice to nail it before I make 5.8 official.
(It still needs work, though. I am upgrading the Guile interface to
the new, improved 1.7 flavor. Not easy. :-( )
Cheers - Bruce
Debian Bug Importer (debzilla) wrote : | #15 |
Message-Id: <email address hidden>
Date: Wed, 30 Nov 2005 04:13:13 -0800
From: Bruce Korb <email address hidden>
To: Matt Kraai <email address hidden>
Cc: Falk Hueffner <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 fixed two license test bugs
On Tuesday 29 November 2005 07:33 pm, Matt Kraai wrote:
> > > I wasn't able to reproduce this problem with autogen 1:5.7.3-1 in the
> > > unstable chroot on escher.debian.org. Falk, would you try to
> > > reproduce the problem and let me know whether or not you are able to?
> >
> > I was able to reproduce it within pbuilder's chroot, but not outside
> > (both sid). Weird. If anybody wants to debug this, I can try to set up
> > an environment where it can be reproduced and provide an account.
>
> If you can give me access to an environment where I can reproduce it,
> I'll investigate further.
Me, too!! :) It would be nice to nail it before I make 5.8 official.
(It still needs work, though. I am upgrading the Guile interface to
the new, improved 1.7 flavor. Not easy. :-( )
Cheers - Bruce
In Debian Bug tracker #340851, Bkorb (bkorb) wrote : Re: Bug#340851: 5.7.3 failure | #16 |
The call to scm_makstr() is faulting (every time for me):
$ uname -a
Linux juist 2.6.13.2 #2 Fri Sep 23 18:45:17 CEST 2005 alpha GNU/Linux
$ gdb ../../autogen
(gdb) set args --trace=everything -b license --no-def -T license.tpl
(gdb) run
Starting program: /home/bkorb/
[Thread debugging using libthread_db enabled]
[New Thread 16384 (LWP 648)]
Starting test template
Text (11) in license.tpl at line 2
/*
EXPR ( B) in license.tpl at line 3
(license "license" "license" "Au
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 648)]
0x0000020000008a4c in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
(gdb) bt
#0 0x0000020000008a4c in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#1 0x0000020000008efc in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#2 0x000002000000c540 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#3 0x000002000000c058 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#4 0x0000000120014654 in ag_scm_license (license=
prog_
at expFormat.c:593
#5 0x0000020000082bd4 in scm_gsubr_apply () from /usr/lib/
#6 0x0000020000068aa4 in scm_dapply () from /usr/lib/
#7 0x0000020000067a90 in scm_deval () from /usr/lib/
#8 0x000002000006395c in scm_i_eval_x () from /usr/lib/
#9 0x000002000006a810 in scm_primitive_
#10 0x0000000120014d0c in ag_scm_
pzExpr=
#11 0x000000012001e224 in ag_eval (
pzStr=
#12 0x0000000120025bd4 in evalExpression (pMustFree=
at funcEval.c:242
#13 0x0000000120026068 in mFunc_Expr (pT=0x1200cf3a0, pMac=0x1200cf428)
at funcEval.c:413
#14 0x00000001200292b4 in generateBlock (pT=0x1200cf3a0, pMac=0x1200cf428,
pEnd=
#15 0x000000012002982c in processTemplate (pTF=0x1200cf3a0) at tpProcess.c:208
#16 0x00000001200076b8 in inner_main (argc=7, argv=0x11ff478f8) at autogen.c:78
#17 0x00000200000790a0 in gh_call3 () from /usr/lib/
#18 0x00000200000873b4 in scm_boot_guile () from /usr/lib/
#19 0x00000200000790e4 in gh_enter () from /usr/lib/
#20 0x00000001200077a4 in main (argc=7, argv=0x11ff478f8) at autogen.c:104
590 /*
591 * Create our output buffer and insert the first prefix
592 */
593 res = scm_makstr( out_size, 0 );
594 pzOut = SCM_CHARS( res );
Debian Bug Importer (debzilla) wrote : | #17 |
Message-Id: <email address hidden>
Date: Thu, 1 Dec 2005 08:11:46 -0800
From: Bruce Korb <email address hidden>
To: Matt Kraai <email address hidden>
Cc: Falk Hueffner <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 failure
The call to scm_makstr() is faulting (every time for me):
$ uname -a
Linux juist 2.6.13.2 #2 Fri Sep 23 18:45:17 CEST 2005 alpha GNU/Linux
$ gdb ../../autogen
(gdb) set args --trace=everything -b license --no-def -T license.tpl
(gdb) run
Starting program: /home/bkorb/
[Thread debugging using libthread_db enabled]
[New Thread 16384 (LWP 648)]
Starting test template
Text (11) in license.tpl at line 2
/*
EXPR ( B) in license.tpl at line 3
(license "license" "license" "Au
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 648)]
0x0000020000008a4c in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
(gdb) bt
#0 0x0000020000008a4c in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#1 0x0000020000008efc in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#2 0x000002000000c540 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#3 0x000002000000c058 in _dl_rtld_di_serinfo () from /lib/ld-linux.so.2
#4 0x0000000120014654 in ag_scm_license (license=
prog_
at expFormat.c:593
#5 0x0000020000082bd4 in scm_gsubr_apply () from /usr/lib/
#6 0x0000020000068aa4 in scm_dapply () from /usr/lib/
#7 0x0000020000067a90 in scm_deval () from /usr/lib/
#8 0x000002000006395c in scm_i_eval_x () from /usr/lib/
#9 0x000002000006a810 in scm_primitive_
#10 0x0000000120014d0c in ag_scm_
pzExpr=
#11 0x000000012001e224 in ag_eval (
pzStr=
#12 0x0000000120025bd4 in evalExpression (pMustFree=
at funcEval.c:242
#13 0x0000000120026068 in mFunc_Expr (pT=0x1200cf3a0, pMac=0x1200cf428)
at funcEval.c:413
#14 0x00000001200292b4 in generateBlock (pT=0x1200cf3a0, pMac=0x1200cf428,
pEnd=
#15 0x000000012002982c in processTemplate (pTF=0x1200cf3a0) at tpProcess.c:208
#16 0x00000001200076b8 in inner_main (argc=7, argv=0x11ff478f8) at autogen.c:78
#17 0x00000200000790a0 in gh_call3 () from /usr/lib/
#18 0x00000200000873b4 in scm_boot_guile () from /usr/lib/
#19 0x00000200000790e4 in gh_enter () from /usr/lib/
#20 0x00000001200077a4 in main (argc=7, argv=0x11ff478f8) at autogen.c:104
590 /*
591 * Create our output buffer and insert the first prefix
592 */
593 res = scm_makstr( out_size, 0 );
594 pzOut = SCM_CHARS( res );
In Debian Bug tracker #340851, Bkorb (bkorb) wrote : | #18 |
On Thursday 01 December 2005 08:11 am, Bruce Korb wrote:
> The call to scm_makstr() is faulting (every time for me):
Hi all,
I have now been able to recreate the failure on this call fairly
often, but I have not found a cause. Nevertheless, this call is
only useful if you use SCM_CHARS() and that has been so strongly
"deprecated" that the current development releases (1.7.x) of
Guile emit warnings when it is used. Rather than fight it, I have
removed all usage of either of these for all variations of AutoGen.
(Each version of Guile causes me to rewrite some portion of
AutoGen. 1.7 is the worst yet, causing me several days of work.)
Upshot: I now (finally) have a 1.7.x-ready version of AutoGen.
Problem: still seg faults on Alpha Linux 2.2. The issue: mmap
Details:
When I mmap a file, I have to test for it being a multiple of the
page size. My "license.test" forces this for both the template
and for the license text. Both of these files get mmapped, so
I use this test to verify proper functioning of mmap, too.
My mapping descriptor structure shows this:
(gdb) print *pMI
$2 = {txt_data = 0x2000001c000, txt_size = 8192, txt_full_size = 16384,
txt_fd = 7, txt_zero_fd = -1, txt_errno = 0, txt_prot = 0, txt_flags = 0,
txt_alloc = 0}
Immediately before this call:
pNuls = mmap(
The address passed is:
(gdb) print 0x2000001c000+8192
$5 = 0x2000001e000
According to my reading of the doc for mmap, this should work. It does
work on other platforms. It also seg faults instead of returning ((void*)-1).
Now, what? :-(
I have a new "pre" release:
http://
that has these changes.
Regards, Bruce
Debian Bug Importer (debzilla) wrote : | #19 |
Message-Id: <email address hidden>
Date: Mon, 5 Dec 2005 13:34:43 -0800
From: Bruce Korb <email address hidden>
To: Matt Kraai <email address hidden>
Cc: Falk Hueffner <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 failure
On Thursday 01 December 2005 08:11 am, Bruce Korb wrote:
> The call to scm_makstr() is faulting (every time for me):
Hi all,
I have now been able to recreate the failure on this call fairly
often, but I have not found a cause. Nevertheless, this call is
only useful if you use SCM_CHARS() and that has been so strongly
"deprecated" that the current development releases (1.7.x) of
Guile emit warnings when it is used. Rather than fight it, I have
removed all usage of either of these for all variations of AutoGen.
(Each version of Guile causes me to rewrite some portion of
AutoGen. 1.7 is the worst yet, causing me several days of work.)
Upshot: I now (finally) have a 1.7.x-ready version of AutoGen.
Problem: still seg faults on Alpha Linux 2.2. The issue: mmap
Details:
When I mmap a file, I have to test for it being a multiple of the
page size. My "license.test" forces this for both the template
and for the license text. Both of these files get mmapped, so
I use this test to verify proper functioning of mmap, too.
My mapping descriptor structure shows this:
(gdb) print *pMI
$2 = {txt_data = 0x2000001c000, txt_size = 8192, txt_full_size = 16384,
txt_fd = 7, txt_zero_fd = -1, txt_errno = 0, txt_prot = 0, txt_flags = 0,
txt_alloc = 0}
Immediately before this call:
pNuls = mmap(
The address passed is:
(gdb) print 0x2000001c000+8192
$5 = 0x2000001e000
According to my reading of the doc for mmap, this should work. It does
work on other platforms. It also seg faults instead of returning ((void*)-1).
Now, what? :-(
I have a new "pre" release:
http://
that has these changes.
Regards, Bruce
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : | #20 |
On Mon, Dec 05, 2005 at 01:34:43PM -0800, Bruce Korb wrote:
> On Thursday 01 December 2005 08:11 am, Bruce Korb wrote:
> > The call to scm_makstr() is faulting (every time for me):
>
> Hi all,
>
> I have now been able to recreate the failure on this call fairly
> often, but I have not found a cause. Nevertheless, this call is
> only useful if you use SCM_CHARS() and that has been so strongly
> "deprecated" that the current development releases (1.7.x) of
> Guile emit warnings when it is used. Rather than fight it, I have
> removed all usage of either of these for all variations of AutoGen.
> (Each version of Guile causes me to rewrite some portion of
> AutoGen. 1.7 is the worst yet, causing me several days of work.)
>
> Upshot: I now (finally) have a 1.7.x-ready version of AutoGen.
> Problem: still seg faults on Alpha Linux 2.2. The issue: mmap
>
> Details:
>
> When I mmap a file, I have to test for it being a multiple of the
> page size. My "license.test" forces this for both the template
> and for the license text. Both of these files get mmapped, so
> I use this test to verify proper functioning of mmap, too.
>
> My mapping descriptor structure shows this:
>
> (gdb) print *pMI
> $2 = {txt_data = 0x2000001c000, txt_size = 8192, txt_full_size = 16384,
> txt_fd = 7, txt_zero_fd = -1, txt_errno = 0, txt_prot = 0, txt_flags = 0,
> txt_alloc = 0}
>
> Immediately before this call:
>
> pNuls = mmap(
> (void*)
> pgsz,
> PROT_READ|
> MAP_ANONYMOUS|
>
> The address passed is:
>
> (gdb) print 0x2000001c000+8192
> $5 = 0x2000001e000
>
> According to my reading of the doc for mmap, this should work. It does
> work on other platforms. It also seg faults instead of returning ((void*)-1).
> Now, what? :-(
If you don't want to require that users use a kernel that doesn't
misbehave like this, shouldn't you add an autoconf test that checks
for this behavior and falls back to some other method?
--
Matt
Debian Bug Importer (debzilla) wrote : | #21 |
Message-ID: <email address hidden>
Date: Tue, 6 Dec 2005 03:50:28 -0800
From: Matt Kraai <email address hidden>
To: Bruce Korb <email address hidden>
Cc: Falk Hueffner <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 failure
--v9Ux+11Zm5mwPlX6
Content-Type: text/plain; charset=us-ascii
Content-
Content-
On Mon, Dec 05, 2005 at 01:34:43PM -0800, Bruce Korb wrote:
> On Thursday 01 December 2005 08:11 am, Bruce Korb wrote:
> > The call to scm_makstr() is faulting (every time for me):
>=20
> Hi all,
>=20
> I have now been able to recreate the failure on this call fairly
> often, but I have not found a cause. Nevertheless, this call is
> only useful if you use SCM_CHARS() and that has been so strongly
> "deprecated" that the current development releases (1.7.x) of
> Guile emit warnings when it is used. Rather than fight it, I have
> removed all usage of either of these for all variations of AutoGen.
> (Each version of Guile causes me to rewrite some portion of
> AutoGen. 1.7 is the worst yet, causing me several days of work.)
>=20
> Upshot: I now (finally) have a 1.7.x-ready version of AutoGen.
> Problem: still seg faults on Alpha Linux 2.2. The issue: mmap
>=20
> Details:
>=20
> When I mmap a file, I have to test for it being a multiple of the
> page size. My "license.test" forces this for both the template
> and for the license text. Both of these files get mmapped, so
> I use this test to verify proper functioning of mmap, too.
>=20
> My mapping descriptor structure shows this:
>=20
> (gdb) print *pMI
> $2 =3D {txt_data =3D 0x2000001c000, txt_size =3D 8192, txt_full_size =3D =
16384,
> txt_fd =3D 7, txt_zero_fd =3D -1, txt_errno =3D 0, txt_prot =3D 0, txt_=
flags =3D 0,
> txt_alloc =3D 0}
>=20
> Immediately before this call:
>=20
> pNuls =3D mmap(
> (void*)
> pgsz,
> PROT_READ|
> MAP_ANONYMOUS|
>=20
> The address passed is:
>=20
> (gdb) print 0x2000001c000+8192
> $5 =3D 0x2000001e000
>=20
> According to my reading of the doc for mmap, this should work. It does
> work on other platforms. It also seg faults instead of returning ((void*=
)-1).
> Now, what? :-(
If you don't want to require that users use a kernel that doesn't
misbehave like this, shouldn't you add an autoconf test that checks
for this behavior and falls back to some other method?
--=20
Matt
--v9Ux+11Zm5mwPlX6
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDlXsEfNd
CTeLQ3wyGxTwhtT
=mj3x
-----END PGP SIGNATURE-----
--v9Ux+
In Debian Bug tracker #340851, Bkorb (bkorb) wrote : | #22 |
Hi Matt,
>>According to my reading of the doc for mmap, this should work. It does
>>work on other platforms. It also seg faults instead of returning ((void*)-1).
>>Now, what? :-(
>
>
> If you don't want to require that users use a kernel that doesn't
> misbehave like this, shouldn't you add an autoconf test that checks
> for this behavior and falls back to some other method?
Yes. Very logical. I was feeling tired, exasperated and lazy all at the
same time. ;) It is not an easy test, of course, so it is easier to
be lazy. I am just so surprised and disappointed to find it failing.
Short term: simply disable mmap for page sized files. Dumb and all,
but it is also what I do for fixincludes. *sigh*. Cheers - Bruce
Debian Bug Importer (debzilla) wrote : | #23 |
Message-ID: <email address hidden>
Date: Tue, 06 Dec 2005 09:38:43 -0800
From: Bruce Korb <email address hidden>
To: Matt Kraai <email address hidden>
CC: Falk Hueffner <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: 5.7.3 failure
Hi Matt,
>>According to my reading of the doc for mmap, this should work. It does
>>work on other platforms. It also seg faults instead of returning ((void*)-1).
>>Now, what? :-(
>
>
> If you don't want to require that users use a kernel that doesn't
> misbehave like this, shouldn't you add an autoconf test that checks
> for this behavior and falls back to some other method?
Yes. Very logical. I was feeling tired, exasperated and lazy all at the
same time. ;) It is not an easy test, of course, so it is easier to
be lazy. I am just so surprised and disappointed to find it failing.
Short term: simply disable mmap for page sized files. Dumb and all,
but it is also what I do for fixincludes. *sigh*. Cheers - Bruce
In Debian Bug tracker #340851, Daniel Schepler (schepler-math) wrote : autogen also FTBFS on i386 now | #24 |
Just a quick note that I've been able to reproduce the failure on the
license.test test on an i386 system as well, so it's apparently not
alpha-specific. I just ran pbuilder on the 1:5.7.3-1 source package twice in
a row, and it failed both times.
--
Daniel Schepler
Debian Bug Importer (debzilla) wrote : | #25 |
Message-Id: <email address hidden>
Date: Mon, 19 Dec 2005 10:35:39 +0100
From: Daniel Schepler <email address hidden>
To: <email address hidden>
Subject: autogen also FTBFS on i386 now
Just a quick note that I've been able to reproduce the failure on the
license.test test on an i386 system as well, so it's apparently not
alpha-specific. I just ran pbuilder on the 1:5.7.3-1 source package twice in
a row, and it failed both times.
--
Daniel Schepler
In Debian Bug tracker #340851, Bruce Korb (bruce-korb) wrote : Re: Bug#340851: autogen also FTBFS on i386 now | #26 |
Daniel Schepler wrote:
>Just a quick note that I've been able to reproduce the failure on the
>license.test test on an i386 system as well, so it's apparently not
>alpha-specific. I just ran pbuilder on the 1:5.7.3-1 source package twice in
>a row, and it failed both times.
>
>
Another quick note: the cause is due to use of a deprecated Guile
interface. By not using scm_makstr() the problem goes away. That
problem is bypassed in 5.8, but the libtool used is broken. The libtool
released last night fixes that issue, but I haven't rolled a new
release yet. 5.8.1 will be the one you want (in a day or two).
It will be just 5.8 with a new libtool.
Regards, Bruce
Debian Bug Importer (debzilla) wrote : | #27 |
Message-ID: <email address hidden>
Date: Mon, 19 Dec 2005 07:30:21 -0800
From: Bruce Korb <email address hidden>
To: Daniel Schepler <email address hidden>, <email address hidden>
Subject: Re: Bug#340851: autogen also FTBFS on i386 now
Daniel Schepler wrote:
>Just a quick note that I've been able to reproduce the failure on the
>license.test test on an i386 system as well, so it's apparently not
>alpha-specific. I just ran pbuilder on the 1:5.7.3-1 source package twice in
>a row, and it failed both times.
>
>
Another quick note: the cause is due to use of a deprecated Guile
interface. By not using scm_makstr() the problem goes away. That
problem is bypassed in 5.8, but the libtool used is broken. The libtool
released last night fixes that issue, but I haven't rolled a new
release yet. 5.8.1 will be the one you want (in a day or two).
It will be just 5.8 with a new libtool.
Regards, Bruce
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : tag 340851 - unreproducible | #28 |
tag 340851 - unreproducible
--
Matt
Debian Bug Importer (debzilla) wrote : | #29 |
Message-ID: <email address hidden>
Date: Thu, 5 Jan 2006 09:24:55 -0800
From: Matt Kraai <email address hidden>
To: <email address hidden>
Subject: tag 340851 - unreproducible
--cWoXeonUoKmBZSoM
Content-Type: text/plain; charset=us-ascii
Content-
Content-
tag 340851 - unreproducible
--=20
Matt
--cWoXeonUoKmBZSoM
Content-Type: application/
Content-
Content-
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDvVZnfNd
byHIDaMiaedukL5
=V5Oz
-----END PGP SIGNATURE-----
--cWoXeonUoKmBZ
In Debian Bug tracker #340851, Matt Kraai (kraai) wrote : Fixed in upload of autogen 1:5.7.3+5.8.1pre4-1 to experimental | #30 |
tag 340851 + fixed-in-
quit
This message was generated automatically in response to an
upload to the experimental distribution. The .changes file follows.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Format: 1.7
Date: Fri, 6 Jan 2006 08:55:33 -0800
Source: autogen
Binary: autogen libopts25-dev libopts25
Architecture: source powerpc
Version: 1:5.7.3+5.8.1pre4-1
Distribution: experimental
Urgency: low
Maintainer: Matt Kraai <email address hidden>
Changed-By: Matt Kraai <email address hidden>
Description:
autogen - an automated text file generator
libopts25 - automated option processing library based on autogen - runtime
libopts25-dev - automated option processing library based on autogen - developmen
Closes: 340851
Changes:
autogen (1:5.7.
.
* New prerelease (closes: Bug#340851).
Files:
1760c13dc00393
8fe339d3b3b6a3
4b4b94cf889ff4
ab3c020f86adc2
d502aa4755e4c5
9cc5378b6739ae
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDvqxEfNd
7uJvYhtfE6YwNRA
=fpWF
-----END PGP SIGNATURE-----
Debian Bug Importer (debzilla) wrote : | #31 |
Message-Id: <email address hidden>
Date: Fri, 06 Jan 2006 10:02:10 -0800
From: Matt Kraai <email address hidden>
To: <email address hidden>
Cc: Matt Kraai <email address hidden>
Subject: Fixed in upload of autogen 1:5.7.3+5.8.1pre4-1 to experimental
tag 340851 + fixed-in-
quit
This message was generated automatically in response to an
upload to the experimental distribution. The .changes file follows.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Format: 1.7
Date: Fri, 6 Jan 2006 08:55:33 -0800
Source: autogen
Binary: autogen libopts25-dev libopts25
Architecture: source powerpc
Version: 1:5.7.3+5.8.1pre4-1
Distribution: experimental
Urgency: low
Maintainer: Matt Kraai <email address hidden>
Changed-By: Matt Kraai <email address hidden>
Description:
autogen - an automated text file generator
libopts25 - automated option processing library based on autogen - runtime
libopts25-dev - automated option processing library based on autogen - developmen
Closes: 340851
Changes:
autogen (1:5.7.
.
* New prerelease (closes: Bug#340851).
Files:
1760c13dc00393
8fe339d3b3b6a3
4b4b94cf889ff4
ab3c020f86adc2
d502aa4755e4c5
9cc5378b6739ae
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDvqxEfNd
7uJvYhtfE6YwNRA
=fpWF
-----END PGP SIGNATURE-----
In Debian Bug tracker #340851, Matt Kraai (kraai) wrote : Bug#340851: fixed in autogen 1:5.8.1-1 | #32 |
Source: autogen
Source-Version: 1:5.8.1-1
We believe that the bug you reported is fixed in the latest version of
autogen, which is due to be installed in the Debian FTP archive:
autogen_
to pool/main/
autogen_5.8.1-1.dsc
to pool/main/
autogen_
to pool/main/
autogen_
to pool/main/
libopts25-
to pool/main/
libopts25_
to pool/main/
A summary of the changes between this version and the previous one is
attached.
Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.
Debian distribution maintenance software
pp.
Matt Kraai <email address hidden> (supplier of updated autogen package)
(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Format: 1.7
Date: Fri, 13 Jan 2006 14:46:53 -0800
Source: autogen
Binary: autogen libopts25-dev libopts25
Architecture: source powerpc
Version: 1:5.8.1-1
Distribution: unstable
Urgency: low
Maintainer: Matt Kraai <email address hidden>
Changed-By: Matt Kraai <email address hidden>
Description:
autogen - an automated text file generator
libopts25 - automated option processing library based on autogen - runtime
libopts25-dev - automated option processing library based on autogen - developmen
Closes: 340851
Changes:
autogen (1:5.8.1-1) unstable; urgency=low
.
* New release (closes: Bug#340851).
Files:
3750010704e06f
db5925cb125c1b
ad4776d1e0d628
99810f82e310fb
07deb280284b76
6562368ecbc183
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFDyEESfNd
DiCfGe17fGMHZLB
=QqZI
-----END PGP SIGNATURE-----
In Debian Bug tracker #340851, Matt Kraai (kraai) wrote : Bug#340851: fixed in autogen 1:5.8.1-3 | #33 |
Source: autogen
Source-Version: 1:5.8.1-3
We believe that the bug you reported is fixed in the latest version of
autogen, which is due to be installed in the Debian FTP archive:
autogen_
to pool/main/
autogen_5.8.1-3.dsc
to pool/main/
autogen_
to pool/main/
libopts25-
to pool/main/
libopts25_
to pool/main/
A summary of the changes between this version and the previous one is
attached.
Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.
Debian distribution maintenance software
pp.
Matt Kraai <email address hidden> (supplier of updated autogen package)
(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Format: 1.7
Date: Wed, 25 Jan 2006 22:49:14 -0800
Source: autogen
Binary: autogen libopts25-dev libopts25
Architecture: source powerpc
Version: 1:5.8.1-3
Distribution: unstable
Urgency: low
Maintainer: Matt Kraai <email address hidden>
Changed-By: Matt Kraai <email address hidden>
Description:
autogen - an automated text file generator
libopts25 - automated option processing library based on autogen - runtime
libopts25-dev - automated option processing library based on autogen - developmen
Closes: 340851 349823
Changes:
autogen (1:5.8.1-3) unstable; urgency=low
.
* Change the timeout in gperf.test to 30 seconds to allow enough time on
m68k.
* Add MAP_PRIVATE to a call to mmap (closes: Bug#340851).
* Change some occurrences of ' to ` in getdefs/opts.def, thanks to
Nicolas François (closes: Bug#349823).
Files:
20d2985e6201f9
b555ea2e4a77c9
f30e7b5580a03a
c58721869db134
27275889252bc2
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFD2HbTfNd
cw8b0S9oZq3eUmX
=Vlrn
-----END PGP SIGNATURE-----
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : reopen 340851 | #34 |
reopen 340851
--
Matt
In Debian Bug tracker #340851, Daniel Schepler (schepler-math) wrote : Still FTBFS under pbuilder | #35 |
I'm still getting FTBFS of autogen under pbuilder with version 1:5.8.1-3 on
i386; from my build log:
...
++ lo_dir=
++ test -d /tmp/buildd/
++ lo_dir=
++ LD_LIBRARY_
+++ echo /tmp/buildd/
++ LIB='/tmp/
++ export LD_LIBRARY_PATH
++ cd testdir
+++ pwd
++ testsubdir=
+++ for f in '${top_builddir}' '${top_srcdir}'
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ for f in '${top_builddir}' '${top_srcdir}'
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ for d in . autoopts agen5
+++ cd /tmp/buildd/
+++ pwd
+++ sort -u
+++ sed 's/^/-I/'
++ dirs='-
-I/tmp/
-I/tmp/
+++ echo -I/tmp/
++ INC='-I/
+++ cd /../bin
+++ pwd
++ f=/bin
++ PATH=/tmp/
++ stdopts=
+++ cd /tmp/buildd/
+++ pwd
++ AG_L='/
++ test_main=yes
++ use_flags=true
++ : '=== Running license.test for license ==='
++ chmod +w license.log
+ echo creating license.tpl
creating license.tpl
+ cat
+ echo creating license.lic
creating license.lic
+ cat
+ cat
+ Csrc=license-extend
+ compile
+ :
+ test Xlicense-extend = X
+ test X = X
+ Cexe=license-extend
+ test X = X
+ Dnam=license
++ echo TEST_TEST_
++ tr '[a-z]-' '[A-Z]_'
+ d=TEST_
+ cmd='gcc -O2 -DTEST_
+ eval gcc -O2 -DTEST_
++ gcc -O2 -DTEST_
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : Re: Bug#340851: Still FTBFS under pbuilder | #36 |
On Fri, Jan 27, 2006 at 11:08:48AM +0100, Daniel Schepler wrote:
> I'm still getting FTBFS of autogen under pbuilder with version 1:5.8.1-3 on
> i386; from my build log:
>
> ...
[...]
> FAIL: license.test
Thanks for reproducing the problem.
> I see the build is somehow succeeding on the buildd's, though... but I don't
> know what's different.
Nor do I. Both upstream and I are investigating.
--
Matt
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : | #37 |
On Fri, Jan 27, 2006 at 11:08:48AM +0100, Daniel Schepler wrote:
> I see the build is somehow succeeding on the buildd's, though... but I don't
> know what's different.
Would you please compile and run the attached program on your system
and let me know what the result is?
--
Matt
In Debian Bug tracker #340851, Daniel Schepler (schepler-math) wrote : | #38 |
Le Vendredi 27 Janvier 2006 16:14, Matt Kraai a écrit :
> On Fri, Jan 27, 2006 at 11:08:48AM +0100, Daniel Schepler wrote:
> > I see the build is somehow succeeding on the buildd's, though... but I
> > don't know what's different.
>
> Would you please compile and run the attached program on your system
> and let me know what the result is?
Running kernel 2.6.15-1-686 (version 2.6.15-3) on my laptop, I got:
daniel@
char at 0xb7fc0000 is 0
did not fault dereferencing 0xB7FC0000
(Although I got several warnings from -Wall.)
--
Daniel Schepler
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : | #39 |
On Fri, Jan 27, 2006 at 04:43:16PM +0100, Daniel Schepler wrote:
> Le Vendredi 27 Janvier 2006 16:14, Matt Kraai a écrit :
> > On Fri, Jan 27, 2006 at 11:08:48AM +0100, Daniel Schepler wrote:
> > > I see the build is somehow succeeding on the buildd's, though... but I
> > > don't know what's different.
> >
> > Would you please compile and run the attached program on your system
> > and let me know what the result is?
>
> Running kernel 2.6.15-1-686 (version 2.6.15-3) on my laptop, I got:
>
> daniel@
> char at 0xb7fc0000 is 0
> did not fault dereferencing 0xB7FC0000
>
> (Although I got several warnings from -Wall.)
Thanks for doing this.
Bruce, is this what you expected?
--
Matt
In Debian Bug tracker #340851, Bkorb (bkorb) wrote : | #40 |
Matt Kraai wrote:
> On Fri, Jan 27, 2006 at 04:43:16PM +0100, Daniel Schepler wrote:
>
>>Le Vendredi 27 Janvier 2006 16:14, Matt Kraai a écrit :
>>
>>>On Fri, Jan 27, 2006 at 11:08:48AM +0100, Daniel Schepler wrote:
>>>
>>>>I see the build is somehow succeeding on the buildd's, though... but I
>>>>don't know what's different.
>>>
>>>Would you please compile and run the attached program on your system
>>>and let me know what the result is?
>>
>>Running kernel 2.6.15-1-686 (version 2.6.15-3) on my laptop, I got:
>>
>>daniel@
>>char at 0xb7fc0000 is 0
>>did not fault dereferencing 0xB7FC0000
>>
>>(Although I got several warnings from -Wall.)
>
> Thanks for doing this.
>
> Bruce, is this what you expected?
Hi Guys,
It is not what I expected, but it is not necessarily invalid.
mmap implementations are described as *trying* to map the data between
between some unmapped pages just so that you can try to extend
the mapping. Given that this program is not doing anything else,
it really *ought* to be the case that the reference faults. There should
be lots of unused virtual address space available.
Cheers - Bruce
P.S. I should not always rely on Open Group docs. I just saw this in
Linux docs:
MAP_ANONYMOUS
The mapping is not backed by any file; the fd and offset argu-
ments are ignored. This flag in conjunction with MAP_SHARED is
which means my original version ought to have always worked, but was
not portable. Well, now with MAP_PRIVATE, it should be portable, too. :)
Anyway, here is some POSIX verbiage that actually implies that the above
reference really should fault (though it is colored "optional"):
The system shall always zero-fill any partial page at the end of an object.
Further, the system shall never write out any modified portions of the last
page of an object which are beyond its end. [MPR] [Option Start]
>> References within the address range starting at pa and continuing for len
>> bytes to whole pages following the end of an object shall result in delivery
>> of a SIGBUS signal. [Option End]
An implementation may generate SIGBUS signals when a reference would cause
an error in the mapped object, such as out-of-space condition.
I suppose I could wrap the anonymous mmap call in sigbus protection, but
holy moly, just how friggin hard should it be to mmap a bloody text file
anyways? Sheesh!!!
In Debian Bug tracker #340851, Matt Kraai (kraai) wrote : Bug#340851: fixed in autogen 1:5.8.1-4 | #41 |
Source: autogen
Source-Version: 1:5.8.1-4
We believe that the bug you reported is fixed in the latest version of
autogen, which is due to be installed in the Debian FTP archive:
autogen_
to pool/main/
autogen_5.8.1-4.dsc
to pool/main/
autogen_
to pool/main/
libopts25-
to pool/main/
libopts25_
to pool/main/
A summary of the changes between this version and the previous one is
attached.
Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.
Debian distribution maintenance software
pp.
Matt Kraai <email address hidden> (supplier of updated autogen package)
(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Format: 1.7
Date: Tue, 31 Jan 2006 18:45:05 -0800
Source: autogen
Binary: autogen libopts25-dev libopts25
Architecture: source powerpc
Version: 1:5.8.1-4
Distribution: unstable
Urgency: low
Maintainer: Matt Kraai <email address hidden>
Changed-By: Matt Kraai <email address hidden>
Description:
autogen - an automated text file generator
libopts25 - automated option processing library based on autogen - runtime
libopts25-dev - automated option processing library based on autogen - developmen
Closes: 340851
Changes:
autogen (1:5.8.1-4) unstable; urgency=low
.
* Add one to pMI->txt_size before passing it to mmap (closes:
Bug#340851).
Files:
f2f288cec0e2f8
4faee8fe22a47e
2da29cdae1abcb
4107c5dcd69351
9768f55b47ac36
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
iD8DBQFD4DAIfNd
fLMoOlg0MbkhYF+
=OoRH
-----END PGP SIGNATURE-----
In Debian Bug tracker #340851, Daniel Schepler (schepler-math) wrote : Re: Bug#340851: Still FTBFS under pbuilder | #42 |
Le Mardi 31 Janvier 2006 17:35, vous avez écrit :
> Daniel Schepler wrote:
> >I got:
> >
> >daniel@
> >Successfully mapped NUL page at 0xb7fe9000 (is 0)
>
> THANK YOU. I now have complete confidence that adding 1 (one) to the
> initial mmap fixes the mmap issues on the alpha.
Umm, just to make it clear, all the results I sent were on my Pentium laptop,
not on alpha. Sorry if I somehow gave the impression these were alpha
results.
--
Daniel Schepler
In Debian Bug tracker #340851, Bruce Korb (bruce-korb) wrote : | #43 |
Daniel Schepler wrote:
>Le Mardi 31 Janvier 2006 17:35, vous avez écrit :
>
>
>>Daniel Schepler wrote:
>>
>>
>>>I got:
>>>
>>>daniel@
>>>Successfully mapped NUL page at 0xb7fe9000 (is 0)
>>>
>>>
>>THANK YOU. I now have complete confidence that adding 1 (one) to the
>>initial mmap fixes the mmap issues on the alpha.
>>
>>
>
>Umm, just to make it clear, all the results I sent were on my Pentium laptop,
>not on alpha. Sorry if I somehow gave the impression these were alpha
>results.
>
>
Oops. Yes, that's right. Nevertheless, alpha had the same issue, only
more reliably. As soon as I have a round tuit, I'll make a check on an
alpha I was just granted access to. Thanks! - Bruce
In Debian Bug tracker #340851, Matt Kraai (kraai-ftbfs) wrote : | #44 |
On Thu, Feb 02, 2006 at 07:01:02AM -0800, Bruce Korb wrote:
> Daniel Schepler wrote:
>
> >Le Mardi 31 Janvier 2006 17:35, vous avez écrit :
> >
> >
> >>Daniel Schepler wrote:
> >>
> >>
> >>>I got:
> >>>
> >>>daniel@
> >>>Successfully mapped NUL page at 0xb7fe9000 (is 0)
> >>>
> >>>
> >>THANK YOU. I now have complete confidence that adding 1 (one) to the
> >>initial mmap fixes the mmap issues on the alpha.
> >>
> >>
> >
> >Umm, just to make it clear, all the results I sent were on my Pentium
> >laptop, not on alpha. Sorry if I somehow gave the impression these were
> >alpha results.
> >
> >
> Oops. Yes, that's right. Nevertheless, alpha had the same issue, only
> more reliably. As soon as I have a round tuit, I'll make a check on an
> alpha I was just granted access to. Thanks! - Bruce
I verified the patch on an Alpha, and 5.8.2 built successfully on
Alpha, so I'm confident the problem's fixed.
--
Matt
In Debian Bug tracker #340851, Daniel Schepler (schepler-math) wrote : | #45 |
Le Vendredi 03 Février 2006 07:06, Matt Kraai a écrit :
> On Thu, Feb 02, 2006 at 07:01:02AM -0800, Bruce Korb wrote:
> > Oops. Yes, that's right. Nevertheless, alpha had the same issue, only
> > more reliably. As soon as I have a round tuit, I'll make a check on an
> > alpha I was just granted access to. Thanks! - Bruce
>
> I verified the patch on an Alpha, and 5.8.2 built successfully on
> Alpha, so I'm confident the problem's fixed.
I can confirm that autogen versions 5.8.1-4 and 5.8.2-1 both built in the
pbuilder environment where it broke before, too.
--
Daniel Schepler
In Debian Bug tracker #340851, Bruce Korb (bruce-korb) wrote : | #46 |
Daniel Schepler wrote:
>Le Mardi 31 Janvier 2006 17:35, vous avez écrit :
>
>
>>Daniel Schepler wrote:
>>
>>
>>>I got:
>>>
>>>daniel@
>>>Successfully mapped NUL page at 0xb7fe9000 (is 0)
>>>
>>>
>>THANK YOU. I now have complete confidence that adding 1 (one) to the
>>initial mmap fixes the mmap issues on the alpha.
>>
>>
>
>Umm, just to make it clear, all the results I sent were on my Pentium laptop,
>not on alpha. Sorry if I somehow gave the impression these were alpha
>results.
>
>
Thanks. I have just verified it on an Alpha, just to be certain.
5.8.3 works on an alpha machine that was failing before. Thanks - Bruce
Automatically imported from Debian bug report #340851 http:// bugs.debian. org/340851