]>
git.ipfire.org Git - thirdparty/valgrind.git/log
Julian Seward [Thu, 10 Jul 2003 00:17:58 +0000 (00:17 +0000)]
Add a new mechanism for intercepting calls, which doesn't depend on
the vagaries of the dynamic linker. In particular this has been
devised so as to work around errno/h_errno/resolver-state misbehaviour
caused by excessive PLT bypassing in glibc-2.3.2: we need to intercept
calls to __errno_location(), __h_errno_location() and __res_state(),
in threaded programs, but we can't always do that because some calls
made internally within glibc-2.3.2 bypass the PLT.
New mechanism is:
- In vg_symtab2.c, VG_(setup_code_redirect_table), search the
symbol tables to find the entry points of the above functions,
and find the corresponding entry points replacements in our
vg_libpthread.c. Put these pairs into a table,
VG_(code_redirect_table).
- In vg_translate.c, VG_(translate), consult the table each time
a translation is made, and if a hit is found, translate from
the substitute address instead.
This seems to make corecheck/tests/res_search work properly,
although for some as-yet unknown reason breaks the corecheck
skin. All other skins appear unaffected.
One unfortunate effect is that the lazy debug info scheme is now
nullified, since we always need to read debug info in order to
generate the redirection table.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1743
Julian Seward [Wed, 9 Jul 2003 21:53:32 +0000 (21:53 +0000)]
Add iffdeffery to allow testing the call to the DNS resolver in either
the main thread, or in some other thread (the failure case).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1742
Nicholas Nethercote [Tue, 8 Jul 2003 20:59:57 +0000 (20:59 +0000)]
Allow for 1 and 2 digit pids
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1741
Julian Seward [Tue, 8 Jul 2003 00:13:51 +0000 (00:13 +0000)]
Dodge around more glibc variants.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1740
Julian Seward [Mon, 7 Jul 2003 23:56:10 +0000 (23:56 +0000)]
Fix a whole lot more regression tests on Red Hat 7.3.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1739
Julian Seward [Mon, 7 Jul 2003 07:58:17 +0000 (07:58 +0000)]
Undo mistaken rev 1.54.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1738
Julian Seward [Mon, 7 Jul 2003 00:43:34 +0000 (00:43 +0000)]
Make more tests glibc-2.X robust.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1737
Julian Seward [Mon, 7 Jul 2003 00:32:44 +0000 (00:32 +0000)]
Try and make different gcc versions agree on source locations. Sigh.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1736
Julian Seward [Mon, 7 Jul 2003 00:23:23 +0000 (00:23 +0000)]
Make some more tests more robust against glibc differences.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1735
Julian Seward [Mon, 7 Jul 2003 00:03:52 +0000 (00:03 +0000)]
A regression test for the new VALGRIND_{GET,SET}_VBITS macros.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1734
Julian Seward [Sun, 6 Jul 2003 23:43:01 +0000 (23:43 +0000)]
Make these work on Red Hat 7.3 (by giving -q)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1733
Julian Seward [Sun, 6 Jul 2003 23:24:18 +0000 (23:24 +0000)]
Run the test with -q so that it works both on glibc-2.2 and 2.3. The
problem if -q is not used is that the lines "malloc/free: 1 allocs, 0
frees, 8 bytes allocated." have different counts depending on the
glibc version.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1732
Julian Seward [Sun, 6 Jul 2003 19:18:42 +0000 (19:18 +0000)]
Make it compile with older compilers.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1731
Dirk Mueller [Sun, 6 Jul 2003 17:01:43 +0000 (17:01 +0000)]
make check more portable, patch by Matthew Emmerton
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1730
Julian Seward [Sun, 6 Jul 2003 01:52:32 +0000 (01:52 +0000)]
Increase default number of supported threads from 50 to 100.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1729
Julian Seward [Sun, 6 Jul 2003 01:40:11 +0000 (01:40 +0000)]
load_one_suppressions_file(): Make sure to finish reading
up-to-and-including the '}' when the number of callers is >=
VG_N_SUPP_CALLERS. (Jeffrey Stedfast)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1728
Julian Seward [Sun, 6 Jul 2003 01:34:13 +0000 (01:34 +0000)]
Restarting valgrind-listener quickly can produce:
cannot bind port : Address already in use
Fix it by enabling SO_REUSEADDR for the listening socket.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1727
Julian Seward [Sun, 6 Jul 2003 01:29:49 +0000 (01:29 +0000)]
Fix up weak binding for pthread_mutexattr_setpshared.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1726
Julian Seward [Sun, 6 Jul 2003 01:23:11 +0000 (01:23 +0000)]
Implement pthread_mutexattr_setpshared (Nick did this).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1724
Julian Seward [Sun, 6 Jul 2003 01:17:33 +0000 (01:17 +0000)]
Handle XCHG eAX,eBP.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1723
Julian Seward [Sun, 6 Jul 2003 01:14:42 +0000 (01:14 +0000)]
string_match_wrk(): increase max recursion depth to 500.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1722
Julian Seward [Sun, 6 Jul 2003 01:11:47 +0000 (01:11 +0000)]
VGR_(select): be robust against negative timeout values (treat as zero).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1721
Julian Seward [Sun, 6 Jul 2003 01:05:10 +0000 (01:05 +0000)]
The Pentium 4 recycles the CS segment override prefix (0x2E) to mean
'predict not taken' when it appears before a conditional branch.
Valgrind 1.9.6 doesn't know how to handle this. The appended patch
makes it ignore the prefix in this case, which should be safe.
A DS segment override means 'predict taken', but valgrind already
ignores that on a conditional branch, so nothing needs to be done.
(Zack Weinberg)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1720
Julian Seward [Sun, 6 Jul 2003 00:54:47 +0000 (00:54 +0000)]
Add patch from Nick to get proper backtraces in
MC_(get_or_set_vbits_for_client).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1719
Julian Seward [Sat, 5 Jul 2003 17:53:55 +0000 (17:53 +0000)]
Add memcheck client requests VALGRIND_GET_VBITS / VALGRIND_SET_VBITS
for fetching/setting metadata so that it can be sent between
unconnected address spaces (or whatever).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1718
Dirk Mueller [Fri, 4 Jul 2003 16:41:22 +0000 (16:41 +0000)]
whops
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1717
Dirk Mueller [Fri, 4 Jul 2003 16:28:58 +0000 (16:28 +0000)]
use the automake feature for dist-bzip2
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1716
Dirk Mueller [Fri, 4 Jul 2003 16:18:15 +0000 (16:18 +0000)]
spelling fixes
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1715
Dirk Mueller [Fri, 4 Jul 2003 16:16:51 +0000 (16:16 +0000)]
add failing test
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1714
Nicholas Nethercote [Fri, 4 Jul 2003 15:59:49 +0000 (15:59 +0000)]
Operator change. No difference in behaviour, but avoids a warning with Perl
5.8.0.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1713
Nicholas Nethercote [Fri, 4 Jul 2003 11:44:39 +0000 (11:44 +0000)]
wibble
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1712
Nicholas Nethercote [Sun, 29 Jun 2003 17:28:23 +0000 (17:28 +0000)]
Added (trivially) "XCHG eAX,eSP" instruction.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1711
Nicholas Nethercote [Sun, 29 Jun 2003 10:18:48 +0000 (10:18 +0000)]
Doubled VG_N_WAITING_FDS from 10 to 20. Joshua Moore-Oliva
<josh@chatgris.com> hit his head on it a while ago.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1710
Nicholas Nethercote [Sun, 29 Jun 2003 10:12:58 +0000 (10:12 +0000)]
Fixed cpuid regtest, which was broken by the switch to using the machine's real
CPUID.
Also added filters to Makefiles where missing; "make regtest" was broken
within distributions created using "make dist", now fixed.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1709
Nicholas Nethercote [Thu, 26 Jun 2003 14:57:01 +0000 (14:57 +0000)]
comment only
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1708
Dirk Mueller [Thu, 19 Jun 2003 18:46:16 +0000 (18:46 +0000)]
make sure that the cpuid of the simulated CPUID doesn't pretend
to support 3dnow!
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1707
Nicholas Nethercote [Thu, 19 Jun 2003 10:14:15 +0000 (10:14 +0000)]
Move two variable declarations to the start of their enclosing blocks.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1706
Julian Seward [Wed, 18 Jun 2003 23:30:52 +0000 (23:30 +0000)]
A few more SSEs.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1705
Julian Seward [Wed, 18 Jun 2003 23:30:14 +0000 (23:30 +0000)]
Make the CPUID helper really do cpuid, so that what is returned
depends on the host CPU. We no longer claim to be Werner Trobin's
P54C or any more recent variant thereof :-)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1704
Julian Seward [Wed, 18 Jun 2003 00:22:01 +0000 (00:22 +0000)]
A few more SSE float (scalar,packed) insns. Something is not right;
the Ogg encoder vectorised for PIII (SSE, -xK) crashes now.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1703
Julian Seward [Mon, 16 Jun 2003 23:34:12 +0000 (23:34 +0000)]
- Add a few more SSE instructions
- Fix a bug in MOVD reg-xmmreg / xmmreg-reg which was causing
programs to misbehave
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1702
Julian Seward [Sun, 15 Jun 2003 23:26:04 +0000 (23:26 +0000)]
Add some more SSE/SSE2 floating point instructions.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1701
Dirk Mueller [Sun, 15 Jun 2003 22:39:15 +0000 (22:39 +0000)]
AC_PROG_EGREP wasn't too portable after all. Better fixing it the lame but
safe way
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1700
Julian Seward [Sun, 15 Jun 2003 22:28:05 +0000 (22:28 +0000)]
- Continue to clear up SSE insn emitters following recent cleanups re
G vs E register fields in encoding
- Add SSE3e_RegWr and use it to implement a case in MOVD.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1699
Julian Seward [Sun, 15 Jun 2003 21:54:34 +0000 (21:54 +0000)]
Rename the SSE and MMX uinstrs which read/write integer registers, in
a way consistent with the position of the register field in the
instruction. In Intel encoding parlance, the G register is in bits
5,4,3 and the E register is bits 2,1,0, and so we adopt this scheme
consistently. Considering how much confusion this has caused me in
this recent bout of SSE hacking, consistent renaming can only be a
good thing. It makes it a lot easier to figure out if parts of the
SSE handling machinery are correct, or not.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1698
Julian Seward [Sun, 15 Jun 2003 11:46:30 +0000 (11:46 +0000)]
Fill in a whole bunch of boring-ass SSE SIMD-integer insns.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1697
Julian Seward [Sun, 15 Jun 2003 03:13:48 +0000 (03:13 +0000)]
Yet more SSE insns.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1696
Julian Seward [Sun, 15 Jun 2003 01:40:58 +0000 (01:40 +0000)]
- Comment out SSE uinstrs which have not so far been used.
- Fix bug in implementation of MOVD (mmxreg <--> ireg). This
fixes various segfaults and strange behaviour.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1695
Julian Seward [Sun, 15 Jun 2003 01:37:27 +0000 (01:37 +0000)]
Remove debug printing that sneaked in.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1694
Julian Seward [Sat, 14 Jun 2003 16:01:32 +0000 (16:01 +0000)]
Teach AddrCheck about some SSE uinstrs.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1693
Julian Seward [Sat, 14 Jun 2003 15:49:25 +0000 (15:49 +0000)]
Do SSE insns PSRLDQ, PSLLDQ
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1692
Julian Seward [Sat, 14 Jun 2003 15:47:15 +0000 (15:47 +0000)]
Get rid of spurious declaration for fooble.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1691
Julian Seward [Sat, 14 Jun 2003 13:22:36 +0000 (13:22 +0000)]
More SSE instructions.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1690
Julian Seward [Sat, 14 Jun 2003 12:05:29 +0000 (12:05 +0000)]
Add yet more SSE/SSE2 instructions. The party never stops here at
Valgrind HQ :-)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1689
Julian Seward [Sat, 14 Jun 2003 12:03:35 +0000 (12:03 +0000)]
- saneUInstr: Allow SSE4 to have CCf (writes ZCP flags).
- vg_improve: add debugging hooks (#if0'd out)
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1688
Julian Seward [Sat, 14 Jun 2003 12:00:45 +0000 (12:00 +0000)]
Fix code generation errors for the SSE3g_ family. This fixes various
segfaults running SSEd code.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1687
Julian Seward [Sat, 14 Jun 2003 11:57:59 +0000 (11:57 +0000)]
When exiting with VgSrc_BbsDone (switching back to real CPU because
block execution count exceeded, debugging only), restore the signal
state before switching back rather than after. I no longer understand
why it had to be done afterwards. This simplifies vg_startup.S a bit.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1686
Nicholas Nethercote [Fri, 13 Jun 2003 15:02:29 +0000 (15:02 +0000)]
several wibbles: added some missing types in function protos, some missing
comments, etc.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1685
Dirk Mueller [Fri, 13 Jun 2003 12:24:41 +0000 (12:24 +0000)]
don't bail out when pthread_once's init route calls itself recursively
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1684
Julian Seward [Fri, 13 Jun 2003 00:26:02 +0000 (00:26 +0000)]
Tonight's installment of SSE madness. Add a couple more nasties
(pextrw/pinsrw).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1683
Nicholas Nethercote [Thu, 12 Jun 2003 20:43:20 +0000 (20:43 +0000)]
Whoops, should have committed this when I added SIOCOUTQ ioctl.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1682
Nicholas Nethercote [Thu, 12 Jun 2003 14:27:58 +0000 (14:27 +0000)]
Added SIOCOUTQ ioctl, thanks to Lennert Buytenhek <buytenh@gnu.org>.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1681
Nicholas Nethercote [Thu, 12 Jun 2003 14:13:05 +0000 (14:13 +0000)]
Some Makefile.am changes. Lots of them affected:
- changed deprecated INCLUDES variable to AM_CPPFLAGS
- moved the -DVG_LIBDIR definition from AM_CFLAGS into AM_CPPFLAGS
- generally neatened them up a bit -- removed old commented out stuff, fixed a
couple of other minor things
Everything works for me, hopefully it won't break things for anyone else...
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1680
Nicholas Nethercote [Thu, 12 Jun 2003 11:24:10 +0000 (11:24 +0000)]
Update, obviously hasn't been for a while, was majorly out of date and
undoubtedly confusing to anyone who read it.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1679
Nicholas Nethercote [Thu, 12 Jun 2003 10:23:58 +0000 (10:23 +0000)]
Minor improvement.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1678
Nicholas Nethercote [Thu, 12 Jun 2003 09:58:41 +0000 (09:58 +0000)]
Added support for Value16 and Addr16 error suppressions.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1677
Julian Seward [Thu, 12 Jun 2003 00:56:06 +0000 (00:56 +0000)]
Fill in a few more SSE/SSE2 insns, with current aim of being able to
run Qt-3.1 as built with "icc -xW" (P4 code generation). Hopefully by
now I've worked through most SSE/SSE2 conceptual nasties, and it's
mostly a question of filling in the gaps.
I think I might have created a bug of some kind with SSE3g_RegWr. My
current test app segfaults if I run without --optimise=no, which makes
me think I've written something erroneous in the UInstr predicates
controlling optimisation. I don't know what though.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1676
Dirk Mueller [Thu, 12 Jun 2003 00:21:15 +0000 (00:21 +0000)]
revert last commit. too unportable
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1675
Dirk Mueller [Wed, 11 Jun 2003 23:29:48 +0000 (23:29 +0000)]
distribute glibc 2.3 suppression file
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1674
Nicholas Nethercote [Tue, 10 Jun 2003 13:39:15 +0000 (13:39 +0000)]
Fixed stupid bugs in overlap detection.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1673
Nicholas Nethercote [Thu, 5 Jun 2003 21:56:00 +0000 (21:56 +0000)]
Fix problem with LAHF.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1672
Nicholas Nethercote [Tue, 3 Jun 2003 14:56:41 +0000 (14:56 +0000)]
Fixed bug in first attempt at 'lahf'. So much for my supposed testing.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1671
Dirk Mueller [Tue, 3 Jun 2003 14:49:54 +0000 (14:49 +0000)]
suppressions
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1670
Nicholas Nethercote [Tue, 3 Jun 2003 14:12:08 +0000 (14:12 +0000)]
wibble
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1669
Nicholas Nethercote [Tue, 3 Jun 2003 13:38:51 +0000 (13:38 +0000)]
Added support for 'lahf', the twin of 'sahf' which was already implemented.
Tested briefly, seems to work.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1668
Nicholas Nethercote [Tue, 3 Jun 2003 12:55:40 +0000 (12:55 +0000)]
Added two syscalls: sendfile64 and rt_sigtimedwait. Had to #include
signal.h in vg_unsafe.h for the latter.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1667
Dirk Mueller [Sun, 1 Jun 2003 23:44:50 +0000 (23:44 +0000)]
fixing minor oddity in configure output
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1666
Dirk Mueller [Sun, 1 Jun 2003 20:05:08 +0000 (20:05 +0000)]
some makefile massaging to avoid that things get compiled during
make install
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1665
Dirk Mueller [Sun, 1 Jun 2003 20:04:10 +0000 (20:04 +0000)]
fix expected output
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1664
Dirk Mueller [Sun, 1 Jun 2003 20:02:30 +0000 (20:02 +0000)]
make the new symbols weak to fix the new_override testcase.
finally no regression test suite failures!
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1663
Dirk Mueller [Sun, 1 Jun 2003 19:10:05 +0000 (19:10 +0000)]
whoops. one more
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1662
Dirk Mueller [Sun, 1 Jun 2003 19:06:03 +0000 (19:06 +0000)]
unbreak make regtest again
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1661
Julian Seward [Sun, 1 Jun 2003 18:49:12 +0000 (18:49 +0000)]
Add some more SSE insns.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1660
Nicholas Nethercote [Fri, 30 May 2003 09:37:26 +0000 (09:37 +0000)]
comments only
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1659
Dirk Mueller [Wed, 28 May 2003 01:02:46 +0000 (01:02 +0000)]
some ignores
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1658
Julian Seward [Tue, 27 May 2003 00:46:28 +0000 (00:46 +0000)]
Fix a design flaw in the SSE decoder. Remove duplication and
generally rationalise it as the structure of Intel's encoding scheme
becomes more apparent. Add a bunch more instructions.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1657
Dirk Mueller [Tue, 27 May 2003 00:21:30 +0000 (00:21 +0000)]
minor code simplification
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1656
Dirk Mueller [Tue, 27 May 2003 00:19:52 +0000 (00:19 +0000)]
add glibc 2.3 suppression file before I loose it again :(
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1655
Dirk Mueller [Tue, 27 May 2003 00:18:49 +0000 (00:18 +0000)]
compiler warning cleanups
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1654
Julian Seward [Mon, 26 May 2003 09:17:41 +0000 (09:17 +0000)]
Teach memcheck about the SSE UInstrs generated thus far. So now the
Qt GL demos run on memcheck.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1653
Julian Seward [Mon, 26 May 2003 08:47:27 +0000 (08:47 +0000)]
Implement enough SSE/SSE2 insns so that all the GL demos in qt-3.1.0
work, when running on a P4 with an NVidia Vanta card and using
NVidia-supplied libGL.so.1.0.3123. Surprisingly this seems to require
only a minimal set of instructions. So far this is only with
--skin=none.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1652
Julian Seward [Sun, 25 May 2003 01:07:34 +0000 (01:07 +0000)]
Implement a few SSE instructions, enough to run bzip2 when compiled
with the Intel C compiler (7.1.009) with vectorisation for SSE engaged
(-xK).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1651
Nicholas Nethercote [Thu, 22 May 2003 16:17:53 +0000 (16:17 +0000)]
wibble
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1650
Nicholas Nethercote [Wed, 21 May 2003 10:48:24 +0000 (10:48 +0000)]
Using __PRETTY_FUNCTION__ and some macros, removed 148 lines of very repetitive
code (also removing the perennial problem of matching the error strings to the
functions they were in).
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1649
Nicholas Nethercote [Wed, 21 May 2003 10:13:39 +0000 (10:13 +0000)]
wibble
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1648
Nicholas Nethercote [Tue, 20 May 2003 18:24:54 +0000 (18:24 +0000)]
Fixed up docs a bit: did the core/skin split properly for suppressions.
Client requests still need to be fixed, though.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1647
Nicholas Nethercote [Tue, 20 May 2003 17:42:26 +0000 (17:42 +0000)]
wibble
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1646
Nicholas Nethercote [Tue, 20 May 2003 17:41:24 +0000 (17:41 +0000)]
Minor addition.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1645
Nicholas Nethercote [Tue, 20 May 2003 17:32:21 +0000 (17:32 +0000)]
Minor docs fixups.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1644
Nicholas Nethercote [Tue, 20 May 2003 16:38:24 +0000 (16:38 +0000)]
Added VALGRIND_MALLOCLIKE_BLOCK and VALGRIND_FREELIKE_BLOCK which allow you to
use a custom-allocator and detect almost as many errors as you could detect if
you used malloc/new/new[]. (eg. leaks detected, free errors, free mismatch,
etc).
Had to fiddle with mac_malloc_wrappers.c a bit to factor out the appropriate
code to be called from the client request handling code. Also had to add a
new element `MAC_AllocCustom' to the MAC_AllocKind type.
Also added a little documentation, and a regression test.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@1643