From: scott snyder <snyder@d0sgif.fnal.gov>
To: egcs@cygnus.com
Subject: egcs check-gcc results on mips-sgi-irix5.3 (970825 snapshot)
Date: Wed, 27 Aug 1997 03:29:01 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
hi -
Here are results from running the gcc.c-torture suite on the 970825
snapshot on a mips-sgi-irix5.3 target.
=== gcc Summary ===
# of expected passes 4874
# of unexpected failures 5
# of expected failures 4
# of unsupported tests 7
Here are the unexpected failures:
FAIL: gcc.c-torture/compile/961203-1.c, -O0
FAIL: gcc.c-torture/compile/961203-1.c, -O1
FAIL: gcc.c-torture/compile/961203-1.c, -O2
FAIL: gcc.c-torture/compile/961203-1.c, -O2 -fomit-frame-pointer -finline-functions
-- 961203-1.c ---------------------------------------------------------------
struct s {
char a[0x32100000];
int x:30, y:30;
};
int
main ()
{
struct s* p;
p = (struct s*) 0;
if (p->x == p->y)
exit (1);
}
-----------------------------------------------------------------------------
These apparently fails due to running out of VM on the system
(there was about ~70M of free swap space when i ran this).
FAIL: gcc.c-torture/execute/921117-1.c execution, -O2 -fomit-frame-pointer -finline-functions
-- 921117-1.c ---------------------------------------------------------------
struct s {
char text[11];
int flag;
} cell;
int
check (struct s p)
{
if (p.flag != 99)
return 1;
return strcmp (p.text, "0123456789");
}
main ()
{
cell.flag = 99;
strcpy (cell.text, "0123456789");
if (check (cell))
abort();
exit (0);
}
-----------------------------------------------------------------------------
When i compile this by hand with the above options, it does indeed abort.
Here is the generated code. It looks like it goofs up when copying
the structure as part of setting up for the call to check(). I've
marked that sequence below.
Compiled with:
gcc -O2 -finline-functions -fomit-frame-pointer -S 921117-1.c
.file 1 "921117-1.c"
.option pic2
# -G value = 0, Cpu = 3000, ISA = 1
# GNU C version egcs-2.90.02 970825 (gcc2-970802 experimental) (mips-sgi-irix5.3) compiled by GNU C version egcs-2.90.02 970825 (gcc2-970802 experimental).
# options passed: -O2 -finline-functions -fomit-frame-pointer
# options enabled: -fdefer-pop -fomit-frame-pointer -fcse-follow-jumps
# -fcse-skip-blocks -fexpensive-optimizations -fthread-jumps
# -fstrength-reduce -fpeephole -fforce-mem -ffunction-cse
# -finline-functions -finline -fkeep-static-consts -fcaller-saves
# -fpcc-struct-return -fdelayed-branch -frerun-cse-after-loop
# -fschedule-insns -fschedule-insns2 -fpic -fexceptions -fsjlj-exceptions
# -fcommon -fverbose-asm -fregmove -falias-check -fargument-alias -mgas
# -mabicalls -meb -mcpu=3000
gcc2_compiled.:
__gnu_compiled_c:
.rdata
.align 2
$LC0:
.byte 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37
.byte 0x38,0x39,0x0
.text
.align 2
.globl check
.ent check
check:
.frame $sp,32,$31 # vars= 0, regs= 2/0, args= 16, extra= 8
.mask 0x90000000,-4
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
subu $sp,$sp,32
.cprestore 16
li $2,99 # 0x00000063
sw $31,28($sp)
sw $28,24($sp)
sw $4,32($sp)
sw $5,36($sp)
sw $6,40($sp)
.set noreorder
.set nomacro
bne $7,$2,$L4
sw $7,44($sp)
.set macro
.set reorder
addu $4,$sp,32
la $5,$LC0
la $25,strcmp
jal $31,$25
j $L5
$L4:
li $2,1 # 0x00000001
$L5:
lw $31,28($sp)
#nop
.set noreorder
.set nomacro
j $31
addu $sp,$sp,32
.set macro
.set reorder
.end check
.align 2
.globl main
.ent main
main:
.frame $sp,48,$31 # vars= 16, regs= 2/0, args= 16, extra= 8
.mask 0x90000000,-4
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
subu $sp,$sp,48
.cprestore 16
la $2,cell+12
li $3,99 # 0x00000063
la $8,$LC0
sw $31,44($sp)
sw $28,40($sp)
sw $3,0($2)
lw $4,0($8)
lw $5,4($8)
lh $6,8($8)
lb $7,10($8)
sw $4,-12($2)
sw $5,-8($2)
sh $6,-4($2)
sb $7,-2($2)
lw $2,0($8) #<-- should be based at cell, not $8.
lw $4,4($8)
lw $5,8($8)
lw $6,12($8)
sw $2,24($sp)
sw $4,28($sp)
sw $5,32($sp)
sw $6,36($sp)
lw $2,36($sp)
#nop
.set noreorder
.set nomacro
beq $2,$3,$L13
addu $4,$sp,24
.set macro
.set reorder
.set noreorder
.set nomacro
j $L12
li $2,1 # 0x00000001
.set macro
.set reorder
$L13:
move $5,$8
la $25,strcmp
jal $31,$25
$L12:
.set noreorder
.set nomacro
beq $2,$0,$L11
move $4,$0
.set macro
.set reorder
la $25,abort
jal $31,$25
$L11:
la $25,exit
jal $31,$25
.end main
.comm cell,16
.globl exit .text
.globl abort .text
.globl strcpy .text
.globl strcmp .text
WARNING: multiple messages have this Message-ID
From: H.J. Lu <hjl@lucon.org>
To: egcs@cygnus.com
Subject: A texinfo patch
Date: Wed, 27 Aug 1997 03:29:01 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
Message-ID: <19970827032901.3pdjjqt_n-QYZ8aTl_SgqCKRy-5hSTuSTDDqqn7ucZ4@z> (raw)
Hi,
I need this patch to do parallel make as well as make maintainer-clean.
--
H.J. Lu (hjl@gnu.ai.mit.edu)
--
Mon Aug 25 17:31:49 1997 H.J. Lu (hjl@gnu.ai.mit.edu)
* util/install-info.c (my_strerror): Define it with
macro as strerror if HAVE_STRERROR is defined.
* Makefile (all): Remove sub-all.
(stmp-sub-all): New target. Renamed from sub-all.
(info, texinfo): Also depend on stmp-sub-all.
(clean): Also remove stmp*.
(maintainer-clean): New target. Same as realclean.
Index: texinfo/Makefile.in
===================================================================
RCS file: /home/work/cvs/gnu/egcs/texinfo/Makefile.in,v
retrieving revision 1.1.1.1
diff -u -r1.1.1.1 Makefile.in
--- Makefile.in 1997/08/26 00:21:29 1.1.1.1
+++ Makefile.in 1997/08/27 03:46:44
@@ -80,7 +80,7 @@
INSTALL='$(INSTALL)' INSTALL_DATA='$(INSTALL_DATA)' \
INSTALL_PROGRAM='$(INSTALL_PROGRAM)'
-all: sub-all texinfo
+all: texinfo
check:
installcheck:
@@ -131,18 +131,26 @@
#configure: configure.in
# cd $(srcdir) && autoconf
-sub-all TAGS:
+TAGS:
for dir in $(SUBDIRS); do \
echo making $@ in $$dir; \
(cd $$dir && $(MAKE) $(MDEFINES) $@ || exit 1); \
done
-.PHONY: sub-all
+
+stmp-sub-all:
+ -rm -f stmp-sub-all
+ for dir in $(SUBDIRS); do \
+ echo making sub-all in $$dir; \
+ (cd $$dir && $(MAKE) $(MDEFINES) sub-all || exit 1); \
+ done
+ touch stmp-sub-all
clean mostlyclean:
for dir in $(SUBDIRS); do \
echo making $@ in $$dir; \
(cd $$dir && $(MAKE) $(MDEFINES) $@ || exit 1); \
done
+ -rm -f stmp*
distclean: clean texclean
for dir in $(SUBDIRS); do \
@@ -155,11 +163,11 @@
$(RM) *.aux *.cp *.cps *.dvi *.fn *.fns *.ky *.kys *.pg *.pgs
$(RM) *.toc *.tp *.tps *.vr *.vrs
-realclean: distclean
+realclean maintainer-clean: distclean
# Let's hope we weren't cross-compiling.
# If we depend on sub-all, this always gets remade. Annoying.
-info texinfo: texinfo.texi
+info texinfo: texinfo.texi stmp-sub-all
./makeinfo/makeinfo$(EXEEXT) -I$(srcdir) texinfo.texi
.PHONY: info
Index: texinfo/util/install-info.c
===================================================================
RCS file: /home/work/cvs/gnu/egcs/texinfo/util/install-info.c,v
retrieving revision 1.1.1.1
diff -u -r1.1.1.1 install-info.c
--- install-info.c 1997/08/26 00:21:32 1.1.1.1
+++ install-info.c 1997/08/26 20:54:26
@@ -39,7 +39,11 @@
char *readfile ();
struct line_data *findlines ();
+#ifdef HAVE_STRERROR
+#define my_strerror strerror
+#else
char *my_strerror ();
+#endif
void fatal ();
void insert_entry_here ();
int compare_section_names ();
@@ -334,6 +338,7 @@
", progname);
}
+#ifndef HAVE_STRERROR
/* Convert an errno value into a string describing the error.
We define this function here rather than using strerror
because not all systems have strerror. */
@@ -349,6 +354,7 @@
return sys_errlist[errnum];
return (char *) "Unknown error";
}
+#endif
\f
/* This table defines all the long-named options, says whether they
use an argument, and maps them into equivalent single-letter options. */
WARNING: multiple messages have this Message-ID
From: Jeffrey A Law <law@hurl.cygnus.com>
To: egcs@cygnus.com
Subject: Re: Here are the g++ test results
Date: Wed, 27 Aug 1997 05:21:33 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
Message-ID: <19970827052133.LvmNOz0_ksPpJfQvwAMOxqFe98SWl6vfiseDAShWvrs@z> (raw)
In-Reply-To: Here are the g++ test results
In message <Pine.PMDF.3.91.970826184842.538979005A-100000@MPS.OHIO-STATE.EDU>
you write:
> # of expected passes 3192
> # of unexpected failures 56
> # of unexpected successes 2
> # of expected failures 95
> # of untested testcases 10
>
> g++.log and gcc.log are available for interested parties by anon. ftp at
> niobium.mps.ohio-state.edu (gzipped). If any other info is needed please
> let me know.
Thanks. I think I see the problem.
It appears that you're picking up include files from /usr/include/g++,
and for whatever reason they are not compatable with the current
compiler.
I believe this is a bug in the dejagnu framework since it should
be providing appropriate overriding -I arguments so that you pick
up include files from the source directory itself rather than the
previously installed include files. To be sure I need a little more
information:
Did you build the entire tree -- in particular, did you build libstdc++
and libio?
What are the contents of your $objdir/libio (just the ls output is
fine).
Jeff
WARNING: multiple messages have this Message-ID
From: Jeffrey A Law <law@hurl.cygnus.com>
To: egcs@cygnus.com
Subject: Re: egcs web page
Date: Wed, 27 Aug 1997 05:35:30 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
Message-ID: <19970827053530.LSKKcx2RsGI30mhiuo6MGE2cGfxmV6T0fW7nv6wiBkY@z> (raw)
In-Reply-To: egcs web page
In message <199708262342.QAA00646@atrus.synopsys.com>you write:
> The web page points to the copyright assignment form, but it seems most
> people who want to hack egcs on an ongoing basis need to do the
> assign.future (sp?) form instead. Right?
Yes, I'd been wondering about this myself.
It almost makes me think we should put a copy of the GNU page which
discusses copyrights on our page -- except that instead of referring
people to text files that are only available on FSF machines we should
make copies of those files available via the web.
For reference:
http://www.gnu.ai.mit.edu/prep/maintain_3.html#SEC3
Discusses copyrights and the various variants of the assignment
forms and when they're appropriate.
jeff
[ Actaully, I would prefer to just link to the GNU page, but
it doesn't have links to the actual forms, which I think we
want as readily available as possible. ]
WARNING: multiple messages have this Message-ID
From: N8TM@aol.com <N8TM@aol.com>
To: egcs@cygnus.com
Subject: libF77
Date: Wed, 27 Aug 1997 05:35:30 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
Message-ID: <19970827053530.f9XYun6AidNCvgl9peeUZ7gZLfQC55Sp1SQ-15GMW-A@z> (raw)
I have tested some improvements for math functions in libF77. For
f__cabs(), z_log(), and z_sqrt(), accuracy is much less dependent on extended
precision. Single precision complex functions as well as double precision
complex SIN() are speeded up by at least a factor of 2 on my PPC. As I have
not been able to get through configure with egcs, I am just offering anyone
interested a look at the code at members.aol.com/n8tm/mathf2c.uu I'd be
interested in including asm() conditional defines for various additional
architectures if someone would test them, as well as any improvements in
accuracy which anyone may find. I have included source code for all basic
transcendental functions where I have found inadequacy in glibc or libmoto.
All of the functions have been tested with -fargument-noalias-global but
might better be modified to use "restrict" appropriately if that is adopted
by gcc. The basic functions have been tested under IRIX and HPUX as well.
I'm hoping with more detailed support for various platforms to get an
improved libF77 eventually.
Tim
WARNING: multiple messages have this Message-ID
From: Jeffrey A Law <law@hurl.cygnus.com>
To: egcs@cygnus.com
Subject: Re: g77 configuring, docs (was Re: 970825 Snapshot Available)
Date: Wed, 27 Aug 1997 05:59:03 -0000 [thread overview]
Message-ID: <m0x3ZBA-0004ecC@ocean.lucon.org> (raw)
Message-ID: <19970827055903.8HjxSlmhARpCuUtGNzybnU0RAU5WMCEaoxg-gl-hUdI@z> (raw)
In-Reply-To: g77 configuring, docs (was Re: 970825 Snapshot Available)
> P.S. I've put a new patch into ~fortran/dev/ (g77-970824.diff.gz)
> that is my first attempt to fix the problem whereby people can't
> build g77 docs from scratch without gcc.
Very cool.
> Instead, any ANSI C compiler should be okay. Jeff is pretty
> busy, and I am too busy still to get into patching egcs myself,
> but maybe Jeff will get some time to apply this soon
I'm tempted to do both a gcc2 -> egcs and f77dev -> egcs merge
as soon as we get the first release out the door. I don't think
we want to introduce too much new code this close to a release.
But, yes, it's very cool stuff.
> he's done a great job so far IMO.
Thanks.
> P.P.S. BTW am I correct in assuming that being able to build
> everything (g77 docs, someday g77, and everything else, but I
> think g77's the only problem child right now?) using an ANSI C,
> not K&R C, compiler is the main concern?
Well, K&R would be even better (some vendors ship bundled compilers
which are K&R only), but being able to build with ANSI C is a big
enough win by itself to make it worth the effort to remove the
GNU-isms.
And we may find that the K&R first stage issues become less and
less of an issue as our build procedures improve and fewer and
fewer companies ship old lame-o K&R compilers with systems and
force users to pay for ANSI compilers.
Jeff
next reply other threads:[~1997-08-27 3:29 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
1997-08-27 3:29 scott snyder [this message]
1997-08-27 3:29 ` A texinfo patch H.J. Lu
1997-08-27 5:21 ` Here are the g++ test results Jeffrey A Law
1997-08-27 5:35 ` egcs web page Jeffrey A Law
1997-08-27 5:35 ` libF77 N8TM
1997-08-27 5:59 ` g77 configuring, docs (was Re: 970825 Snapshot Available) Jeffrey A Law
1997-09-10 12:02 ` A texinfo patch Jim Wilson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=m0x3ZBA-0004ecC@ocean.lucon.org \
--to=snyder@d0sgif.fnal.gov \
--cc=egcs@cygnus.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).