May 1999 - XEmacs-Beta - XEmacs Mailing List Archives

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

SL Baur

Andreas Jaeger <aj(a)arthur.rhein-neckar.de> writes in xemacs-beta(a)xemacs.org: >>>>>> Jan Vroonhof writes: ... Jan> Yes, and therefore XEmacs uses glibc 2.x's implementation of it instead Jan> of using its own. That is different in FSF Emacs isn't it? > No. FSF Emacs 20.3 uses the same logic as XEmacs does (the patches > have been ported from FSF Emacs to XEmacs) - and that confuses me > since it breaks on FSF Emacs but works without problems on my system. No, it doesn't. I copied most of the code, but I did not copy the part that unconditionally disabled mmap for allocation. Instead, I forced mmap allocation off before dump, when it is absolutely unsafe, and I forced the use of minimal tag bits so that it could be used after dump. This code has had very heavy testing by me and many others for 15 months now. I am fairly confident about its stability. > I guess we can ignore the patch for XEmacs - as long as nobody > encounters this problem with XEmacs. We can, I suppose. voodoo_free_hook is ugly, but it works. What we really want is to not dump at all. What's the current status of getting rid of dumping? -- I protest the NATO war in Yugoslavia.

1 participants
0 comments

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

SL Baur

[I think my messages are getting out now.] Jan Vroonhof <vroonhof(a)math.ethz.ch> writes in xemacs-beta(a)xemacs.org: > [Ok, ignore all my previous ramblings....I checked FSF 20.3's source now.] > Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: >> Yes, and therefore XEmacs uses glibc 2.x's implementation of it instead >> of using its own. That is different in FSF Emacs isn't it? What is different from FSF Emacs is that XEmacs allows usage of mmap for allocation *after* dump. I checked the FSF Emacs source very carefully and the version I looked at had it always disabled, by apparent accident. When mmap is used for allocation with the Doug Lea malloc, the addresses returned can conflict with Emacs tag bits in the high-bits range of the address, hence ... > So FSF does use the glibc allocator. However Steve forced XEmacs > minimal tag bit support to on when using the Doug lea allocator. I > cannot remember why. Because the tag bits stomp on part of the return address pointer *when mmap for allocation is enabled*. FSF Emacs has to turn mmap off because they don't have the option of minimal tag bits. > FSF Emacs doesn't have minimal tag bit support doesn't it? That > could be the difference but it seems to me now to be more pure luck. It was not luck, except the fortunate circumstance that Kyle implemented it around the time I ready was to use it. > (XEmacs also doesn't free the malloc_state_ptr but surely that cannot > be it. Is that a problem? We don't free *anything* starting just before exit(3) is called (the voodoo_free_hook hack). This is due to the way egcs/gcc-2.9 has unconditional support for constructors & destructors. Bypassing free(3) at exit(3) time unconditionally also bypasses all destructor calls. I made a test on Demeter last night before she died, and I could not reproduce the exit(3) crash described above, but that may be due to differences between glibc-2.1 and glibc-2.0.7preX, or something else. I have upgraded practically everything between March of 1998 and now. -- I protest the NATO war in Yugoslavia.

1 participants
0 comments

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

Jan Vroonhof

[Ok, ignore all my previous ramblings....I checked FSF 20.3's source now.] Jan Vroonhof <vroonhof(a)math.ethz.ch> writes: > Yes, and therefore XEmacs uses glibc 2.x's implementation of it instead > of using its own. That is different in FSF Emacs isn't it? So FSF does use the glibc allocator. However Steve forced XEmacs minimal tag bit support to on when using the Doug lea allocator. I cannot remember why. FSF Emacs doesn't have minimal tag bit support doesn't it? That could be the difference but it seems to me now to be more pure luck. (XEmacs also doesn't free the malloc_state_ptr but surely that cannot be it. > It isn't. It is just annoying because these fake implementation make > XEmacs miscompile. Just for the record: That is in an XEmacs bug. Jan

1 participants
0 comments

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

Andreas Jaeger

>>>>> Jan Vroonhof writes: Jan> Andreas Jaeger <aj(a)arthur.rhein-neckar.de> writes: >> I'm confused. The system malloc in glibc 2.x *is* doug lea malloc. >> I helped getting it working for XEmacs. Jan> Yes, and therefore XEmacs uses glibc 2.x's implementation of it instead Jan> of using its own. That is different in FSF Emacs isn't it? No. FSF Emacs 20.3 uses the same logic as XEmacs does (the patches have been ported from FSF Emacs to XEmacs) - and that confuses me since it breaks on FSF Emacs but works without problems on my system. I guess we can ignore the patch for XEmacs - as long as nobody encounters this problem with XEmacs. Jan> P.S. Why does Glibc 2.1.x define I_SETSIG ? >> AFAIK because it's part of the required interface of <stropts.h> - but >> all stream functions set errno to ENOSYS. If you think this is a bug, >> send a bugreport via glibcbug or the www interface at >> <http://www-gnats.gnu.org:8080/cgi-bin/wwwgnats.pl>. Jan> It isn't. It is just annoying because these fake implementation make Jan> XEmacs miscompile. :-( Andreas -- Andreas Jaeger aj(a)arthur.rhein-neckar.de jaeger(a)informatik.uni-kl.de for pgp-key finger ajaeger(a)aixd1.rhrk.uni-kl.de

1 participants
0 comments

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

Jan Vroonhof

Andreas Jaeger <aj(a)arthur.rhein-neckar.de> writes: > I'm confused. The system malloc in glibc 2.x *is* doug lea malloc. > I helped getting it working for XEmacs. Yes, and therefore XEmacs uses glibc 2.x's implementation of it instead of using its own. That is different in FSF Emacs isn't it? > Jan> P.S. Why does Glibc 2.1.x define I_SETSIG ? > AFAIK because it's part of the required interface of <stropts.h> - but > all stream functions set errno to ENOSYS. If you think this is a bug, > send a bugreport via glibcbug or the www interface at > <http://www-gnats.gnu.org:8080/cgi-bin/wwwgnats.pl>. It isn't. It is just annoying because these fake implementation make XEmacs miscompile. Jan

1 participants
0 comments

Assertion failure int event-msw.c 25 years, 7 months

David Bush

On repeated exposure events in the Cygwin build I get an assertion failure in mswindows_dequeue_dispatch_event. I've never looked at this code before so there is little chance I could resolve the issue by the deadline for the next beta. If no one else can get a quick handle on the problem I will look more closely fore the following beta. I found it easy to reproduce the problem in gnus by simply getting into a newsgroup and repeatedly covering part of the Xemacs frame and then minimizing the covering window.

4 participants
7 comments

What to do about package-index signatures? 25 years, 7 months

Jan Vroonhof

[This message of sheer frustration is dedicated to PGP inc and the US goverment] I really don't know. I am tempting to suggest switching it off[1]. With the new patch one can use mailcryptng to verify the signature, but making sure that a failure realy indicates a fake index seems close to impossible at the moment. A prompt "Package signature validation failed. Continue?" seems like asking for trouble to me. Also, all these prompts can come at the weirdest moments (since the package-get code gets invoked behind the users back sometimes). sigh. Jan Footnotes: [1] At least for 21.1.x until after a rethink.

1 participants
0 comments

Problem with GPG signature format 25 years, 7 months

Oscar Figueiredo

-----BEGIN PGP SIGNED MESSAGE----- I can't use `pui-list-packages' anymore because of the new signature format, here's the top of my *Mailcrypt* buffer: Pretty Good Privacy(tm) 2.6.3i - Public-key encryption for the masses. (c) 1990-96 Philip Zimmermann, Phil's Pretty Good Software. 1996-01-18 International version - not for use in the USA. Does not use RSAREF. Current time: 1999/05/31 16:03 GMT ERROR: Header line added to ASCII armor: "Hash: SHA1" ASCII armor corrupted. I do not complain about the new signature format itself, maybe I should upgrade just my PGP to some version that groks this new "Hash" line but I don't want to right now and I just wish I could disable all that signature verification stuff. I want to live dangerously with no package signatures. Oscar -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: noconv Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface iQBVAwUBN1K0NZ7CaiYEgqPZAQEd1AIAxgA4NHm/FNYK4SJjssAlk32rTX7FsngQ 6ZIKHfxif6+zeHRc3nA2Bu6IU0U7IC0f6bdYP4uVaDLlwbKVTp+H1Q== =G70y -----END PGP SIGNATURE-----

1 participants
0 comments

Re: [Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>] more on the emacs- 25 years, 7 months

Andreas Jaeger

Jan> Andreas Jaeger <aj(a)arthur.rhein-neckar.de> writes: >> I thought the following might be of interest to you. We (glibc >> developers) received a bug report about emacs-20.3 with glibc 2.1.1. >> I can't reproduce this with XEmacs 21.0p2. Jan> That's probably 21.1p2. The reason you don't see it is that on Jan> glibc2.x systems XEmacs uses the system malloc instead of its own doug Jan> lea allocator. The main reason for this is that we don't run into Jan> obscure and hard to debug problems like this. I'm confused. The system malloc in glibc 2.x *is* doug lea malloc. I helped getting it working for XEmacs. Nevertheless I've seen another message that the fix I've send is not optimal (should crash if MALLOC_CHECK_ is set) and therefore shouldn't be applied. I'll keep you updated. Jan > From glancing at the patch in its current incarnation, Jan> it seems awfully glibc specific to me. If I understand it correctly. Jan> It works because the glibc startup code works by calling this hook Jan> somewhere. That's right. Jan> Who says the Solaris startup code will do that? Solaris doesn't use doug lea malloc - or has somebody added general support for it in XEmacs in the meantime? Jan> P.S. Why does Glibc 2.1.x define I_SETSIG ? AFAIK because it's part of the required interface of <stropts.h> - but all stream functions set errno to ENOSYS. If you think this is a bug, send a bugreport via glibcbug or the www interface at <http://www-gnats.gnu.org:8080/cgi-bin/wwwgnats.pl>. Andreas -- Andreas Jaeger aj(a)arthur.rhein-neckar.de jaeger(a)informatik.uni-kl.de for pgp-key finger ajaeger(a)aixd1.rhrk.uni-kl.de

1 participants
0 comments

A few problems with XEmacs 21.2.11. 25 years, 7 months

Dan Glastonbury

G'Day, I'm having a problem with package management and Shell-script mode with font-lock-mode. Firstly to the Shell-script mode problem. If I load a shell script, say my .bashrc file, and then try turning on font-locking, the following message appears in the mini buffer: No match 2 in highlight (2 font-lock-comment-face t) This occurs if I start xemacs with xemacs, xemacs -vanilla, xemacs -q -no-site-file. My other problem is when I try required packages in the package list buffer with the `r' key. I get in the mini-buffer" Wrong type argument: number-char-or-marker-p, nil How do I fix these problems?! I'm running XEmacs 21.2.11 on Windows NT/SP4 (XEmacs 21.2 "Calliope" (cygwin32) of Tue Mar 02 1999 on polyester) with version 1.08 of sh-script package and thanks in advance Dan --KAA00789.921540713/gauss.adelaide.maptek.com.au Content-Type: text/plain -- Daniel Glastonbury <dan at adelaide.maptek.com.au> --------------------------------------------------- It is so hot in some parts of the world that the people there have to live other places.

2 participants
1 comments

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

XEmacs-Beta May 1999