8:21 p.m.
Stranger and stranger....
Based on Stephen suggestions, I took the -g out of cflags. With -O it
still failed in faces. Then I set cflags to +O3 -- and it worked !!!?
So, basically on hpux 10.20 with HP's ansic compiler,
what `which cc`
/usr/bin/cc:
LINT A.10.32.22 CXREF A.10.32.22
HP92453-01 A.10.32.22 HP C Compiler
/usr/lib/libc: $Revision: 76.3 $
You can build xemacs 21.4.6 only with the +O3 or +O4 optimization
levels. Debug, or even default optimization will result in a temacs
which fails at loading and dumping faces.el.
Only problem now is that I was trying to build with debug to see if I
could trace down the hangs I'm getting apparently when kill-new tells
interprogram-cut-function to use own-clipboard. Has anyone else seen
this hang?
gary
9:18 p.m.
New subject: Really need help building 21.4.6 (faces.el fails) 2nd request
On Fri, 05 Apr 2002 14:21:04 EST, Gary Beckmann <gbeckmann(a)RADIONICS.com> said:
You can build xemacs 21.4.6 only with the +O3 or +O4 optimization
levels. Debug, or even default optimization will result in a temacs
which fails at loading and dumping faces.el.
That's just so... *wrong* somehow.
Usually, you have to turn OFF optimization to prevent the compiler from
screwing you to the wall. The only thing I can think of is that either
your compiler is on serious crack when run at low optimizations, or that
the linker/loader can't deal with the large debug symbol tables created
when -g is present. However, you'd expect the symbol table size to be
fairly invariant between +O and +O4.
Anybody want to add this one to the PROBLEMS file? ;)
--
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech
12:21 a.m.
New subject: Really need help building 21.4.6 (faces.el fails) 2nd request
Gary Beckmann <gbeckmann(a)RADIONICS.com> wrote:
So, basically on hpux 10.20 with HP's ansic compiler,
> what `which cc`
/usr/bin/cc:
LINT A.10.32.22 CXREF A.10.32.22
HP92453-01 A.10.32.22 HP C Compiler
/usr/lib/libc: $Revision: 76.3 $
Two things:
1. Upgrade your compiler. I don't know how old this is, but it's old.
If it's not too old (although I suspect it is, as I haven't had any
HP compiler -g problems in years and YEARS), you can probably
download a patch. If it's really old, you may have to buy the
upgrade.
2. Looking at /usr/bin/cc won't necessarily give you useful information,
as it's just a driver program. You need to look at the main compiler
engine -- for example:
$ strings `which cc` | grep ccom
/opt/ansic/lbin/ccom
/ccom
$ what /opt/ansic/lbin/ccom
/opt/ansic/lbin/ccom:
LINT A.10.32.30 CXREF A.10.32.30
HP92453-01 A.10.32.30 HP C Compiler
HP-UX SLLIC/OPTIMIZER UX.10.20.565 (DART45): 03/19/99
Ucode Code Generator - UX10.20.70 (DART45 IZ3) (PACG_UX10)
REV: HP SESD Code
High Level Optimizer - UX.10.21.990303 (UX10.MULTI) [-DHLO_RELEASE +O3] -
01-Apr-99.12:31
/usr/lib/libc: $Revision: 76.3 $
[ Wow. 1999 -- and I've got a pretty "up-to-date" compiler, if not
the absolute latest for HP-UX 10.20. ;-( ]
--
Darryl Okahata
darrylo(a)soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
1:36 p.m.
New subject: Really need help building 21.4.6 (faces.el fails) 2nd request
Darryl,
My compiler is only two months behind yours...
: gary＠rules[14] ; strings `which cc` | grep ccom
/opt/ansic/lbin/ccom
/ccom
: gary＠rules[15] ; what /opt/ansic/lbin/ccom
/opt/ansic/lbin/ccom:
LINT A.10.32.22 CXREF A.10.32.22
HP92453-01 A.10.32.22 HP C Compiler
HP-UX SLLIC/OPTIMIZER UX.10.20.563 (DAVIS): 01/28/99
Ucode Code Generator - UX10.20.67 (PACG_UX10)
REV: HP SESD Code
High Level Optimizer - UX.10.21.980511 (UX10.MULTI) [-DHLO_RELEASE +O3] -
05-Feb-99.05:21
/usr/lib/libc: $Revision: 76.3 $
Upgrading the compiler is not an option (you have no idea of the
amount of regression testing that would cause). Since 10.20 is a dead
branch of HP-UX, there is no argument I could bring to convince folks
here to do a compiler upgrade. Running optimized is fine. I just
won't get my line numbers....
Besides, I believe I have solved the hangs I was looking for. Using
customize (which I discovered had a category "killing"), I have turned
off the use of the clipboard, and haven't had any hangs yet. I'm
still waiting a few days to confirm that before I "officially" report
that as a work-around (at least for HP-UX 10.20, if you experience the
hangs). I don't seem to need the clipboard anyway. The primary
selection is sufficient, and is what users here are used to for all
the other inter-process transfers.
gary
8066
days inactive
8069
days old
3 comments
3 participants
participants (3)
-
Darryl Okahata -
Gary Beckmann -
Valdis Kl=?iso8859-4?Q?=BA?=tnieks