]> git.ipfire.org Git - thirdparty/valgrind.git/log
thirdparty/valgrind.git
21 years agoYet another possible output
Julian Seward [Mon, 23 May 2005 14:15:56 +0000 (14:15 +0000)] 
Yet another possible output

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3786

21 years agoMore possible outputs (sigh)
Julian Seward [Mon, 23 May 2005 14:13:34 +0000 (14:13 +0000)] 
More possible outputs (sigh)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3785

21 years agoThis seems to have been missing.
Julian Seward [Mon, 23 May 2005 14:09:52 +0000 (14:09 +0000)] 
This seems to have been missing.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3784

21 years agoAdd another possible output.
Julian Seward [Mon, 23 May 2005 14:02:51 +0000 (14:02 +0000)] 
Add another possible output.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3783

21 years agoAddress item in NOTES.txt.
Nicholas Nethercote [Sat, 21 May 2005 19:49:09 +0000 (19:49 +0000)] 
Address item in NOTES.txt.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3782

21 years agoConcise-ify switch.
Nicholas Nethercote [Sat, 21 May 2005 19:36:45 +0000 (19:36 +0000)] 
Concise-ify switch.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3781

21 years agoAdd a folding rule for 1Uto64.
Julian Seward [Fri, 20 May 2005 19:21:45 +0000 (19:21 +0000)] 
Add a folding rule for 1Uto64.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1202

21 years agoDither
Julian Seward [Fri, 20 May 2005 02:45:15 +0000 (02:45 +0000)] 
Dither

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3780

21 years agoPrint error counts in the XML output.
Julian Seward [Fri, 20 May 2005 02:29:52 +0000 (02:29 +0000)] 
Print error counts in the XML output.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3779

21 years agoApparently a CFI record covering 1.4MB was spotted in the wild. I
Julian Seward [Fri, 20 May 2005 02:10:45 +0000 (02:10 +0000)] 
Apparently a CFI record covering 1.4MB was spotted in the wild.  I
don't know if I believe this, but anyway, increase the sanity
threshold length accordingly.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3778

21 years agofix trivial regtest failures
Julian Seward [Fri, 20 May 2005 02:02:10 +0000 (02:02 +0000)] 
fix trivial regtest failures

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3777

21 years agoSmall but critical performance fix from Pete Moceyunas: give
Julian Seward [Thu, 19 May 2005 10:54:01 +0000 (10:54 +0000)] 
Small but critical performance fix from Pete Moceyunas: give
each small allocation size its own list.  Otherwise there can
be very long searches along lists looking for a block of the
right size.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3776

21 years agoDon't try and set up the thread state members that have been removed
Tom Hughes [Thu, 19 May 2005 07:32:05 +0000 (07:32 +0000)] 
Don't try and set up the thread state members that have been removed
on amd64 systems.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3775

21 years agoRemove dead field.
Nicholas Nethercote [Thu, 19 May 2005 00:36:20 +0000 (00:36 +0000)] 
Remove dead field.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3774

21 years agoFirst pass at adding ability for Memcheck to print all output in XML
Julian Seward [Thu, 19 May 2005 00:25:45 +0000 (00:25 +0000)] 
First pass at adding ability for Memcheck to print all output in XML
form.  The relevant flag is --xml=yes.  Currently this only works with
Memcheck.

Specifying this flag fixes various other options relating to verbosity
and behaviour of the leak checker, so that the resulting output is in
a relatively fixed form suitable for parsing by GUIs.

Still to do:

* Add mechanism to show error counts
* Add regression test
* Document the resulting format

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3773

21 years agoRemoved the clone_flags, parent_tidptr and child_tidptr flags, which were
Nicholas Nethercote [Wed, 18 May 2005 22:56:00 +0000 (22:56 +0000)] 
Removed the clone_flags, parent_tidptr and child_tidptr flags, which were
written but never read, from os_thread_t.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3772

21 years agoComment only change: improve comments on VG_(clone) for x86 and AMD64.
Nicholas Nethercote [Wed, 18 May 2005 22:42:24 +0000 (22:42 +0000)] 
Comment only change:  improve comments on VG_(clone) for x86 and AMD64.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3771

21 years agoMake VG_(wait_for_threadstate) local to linux/core_os.c.
Nicholas Nethercote [Wed, 18 May 2005 19:50:20 +0000 (19:50 +0000)] 
Make VG_(wait_for_threadstate) local to linux/core_os.c.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3770

21 years agoMake VG_(sanitize_client_sigmask)() local to m_syscalls.
Nicholas Nethercote [Wed, 18 May 2005 19:47:46 +0000 (19:47 +0000)] 
Make VG_(sanitize_client_sigmask)() local to m_syscalls.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3769

21 years agoUn-export VG_(is_sig_ign)().
Nicholas Nethercote [Wed, 18 May 2005 19:43:09 +0000 (19:43 +0000)] 
Un-export VG_(is_sig_ign)().

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3768

21 years agoHandle XCHG Gb,Eb.
Julian Seward [Wed, 18 May 2005 12:04:04 +0000 (12:04 +0000)] 
Handle XCHG Gb,Eb.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1201

21 years agoHandle XCHG rAX, reg for 32-bit regs as well as 64-bit regs. I'm not
Julian Seward [Wed, 18 May 2005 11:47:47 +0000 (11:47 +0000)] 
Handle XCHG rAX, reg for 32-bit regs as well as 64-bit regs.  I'm not
sure this is right -- the AMD64 docs are very difficult to interpret
on the subtle point of precisely what is and isn't to be regarded as a
no-op.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1200

21 years agoHandle XOR Ib, AL.
Julian Seward [Wed, 18 May 2005 10:22:47 +0000 (10:22 +0000)] 
Handle XOR Ib, AL.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1199

21 years agowibble
Nicholas Nethercote [Wed, 18 May 2005 04:52:17 +0000 (04:52 +0000)] 
wibble

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3767

21 years agoHmm, actually put some code into m_options.c.
Nicholas Nethercote [Wed, 18 May 2005 04:36:01 +0000 (04:36 +0000)] 
Hmm, actually put some code into m_options.c.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3766

21 years agoDon't need to -I $(builddir)/include now that valgrind.h isn't
Nicholas Nethercote [Tue, 17 May 2005 21:14:54 +0000 (21:14 +0000)] 
Don't need to -I $(builddir)/include now that valgrind.h isn't
generated from valgrind.h.in.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3765

21 years agominor cleanups
Nicholas Nethercote [Tue, 17 May 2005 21:10:11 +0000 (21:10 +0000)] 
minor cleanups

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3764

21 years agoAvoid compiler warnings.
Nicholas Nethercote [Tue, 17 May 2005 21:07:46 +0000 (21:07 +0000)] 
Avoid compiler warnings.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3763

21 years agoAdd missing #include.
Nicholas Nethercote [Tue, 17 May 2005 21:02:03 +0000 (21:02 +0000)] 
Add missing #include.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3762

21 years agoPass an explicit "-m32" or "-m64" in the C flags on x86/amd64 systems
Tom Hughes [Tue, 17 May 2005 16:13:15 +0000 (16:13 +0000)] 
Pass an explicit "-m32" or "-m64" in the C flags on x86/amd64 systems
so that we get the right compiler when --host is used to ask for an x86
build on amd64 or vice versa.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3761

21 years agoGet rid of all the duplicated settings of AM_CFLAGS in the makefiles
Tom Hughes [Tue, 17 May 2005 16:11:40 +0000 (16:11 +0000)] 
Get rid of all the duplicated settings of AM_CFLAGS in the makefiles
for the various core modules and use a single global setting so that
the flags are actually consistent everywhere.

Also get rid of most the USE_PIE blocks in the makefiles by having
configure pass expand a variable in the new global AM_CFLAGS to one
of "-fpie" or "" depending on whether PIE is in use.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3760

21 years agoGive this file a standard preamble, since having empty C files is
Julian Seward [Tue, 17 May 2005 13:45:04 +0000 (13:45 +0000)] 
Give this file a standard preamble, since having empty C files is
thoroughly confusing.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3759

21 years agocomment-only wibble
Julian Seward [Tue, 17 May 2005 13:43:54 +0000 (13:43 +0000)] 
comment-only wibble

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3758

21 years agoCorrect file name in error messages.
Tom Hughes [Tue, 17 May 2005 11:03:52 +0000 (11:03 +0000)] 
Correct file name in error messages.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3757

21 years agoRemove orphaned comment.
Nicholas Nethercote [Tue, 17 May 2005 04:28:23 +0000 (04:28 +0000)] 
Remove orphaned comment.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3756

21 years agoMove linux-only syscall wrappers into m_syscalls.
Nicholas Nethercote [Tue, 17 May 2005 04:20:31 +0000 (04:20 +0000)] 
Move linux-only syscall wrappers into m_syscalls.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3755

21 years agoUndo the awful Leak Error type-abuse.
Nicholas Nethercote [Tue, 17 May 2005 04:00:11 +0000 (04:00 +0000)] 
Undo the awful Leak Error type-abuse.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3754

21 years agoUpdate website address.
Nicholas Nethercote [Tue, 17 May 2005 03:22:38 +0000 (03:22 +0000)] 
Update website address.

MERGE TO 2.4 REPOSITORY

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3753

21 years agoA 4999 entry hashtable is too small when there are a huge number (500k
Julian Seward [Tue, 17 May 2005 02:20:35 +0000 (02:20 +0000)] 
A 4999 entry hashtable is too small when there are a huge number (500k
ish) of blocks simultaneously live, and the client is doing intensive
malloc/frees.  Increase the table size 16-fold.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3752

21 years agoAdded module m_options for holding all the command-line option stuff.
Nicholas Nethercote [Mon, 16 May 2005 23:31:24 +0000 (23:31 +0000)] 
Added module m_options for holding all the command-line option stuff.
Perhaps parts of process_cmd_line_option() should go in here, but I've
not done that for now.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3751

21 years agoFix test for amd64.
Tom Hughes [Mon, 16 May 2005 23:16:25 +0000 (23:16 +0000)] 
Fix test for amd64.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3750

21 years agoMove VGA_STACK_SIZE_W into m_syscalls, the only place that uses it.
Nicholas Nethercote [Mon, 16 May 2005 21:39:23 +0000 (21:39 +0000)] 
Move VGA_STACK_SIZE_W into m_syscalls, the only place that uses it.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3749

21 years agoRemoved the core_arch_asm.h files. The OYNK macros now live in tool_asm.h,
Nicholas Nethercote [Mon, 16 May 2005 21:30:24 +0000 (21:30 +0000)] 
Removed the core_arch_asm.h files.  The OYNK macros now live in tool_asm.h,
at least for the moment.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3748

21 years agoMove ThreadArchState into core.h and avoid duplication.
Nicholas Nethercote [Mon, 16 May 2005 21:16:44 +0000 (21:16 +0000)] 
Move ThreadArchState into core.h and avoid duplication.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3747

21 years agoRemove VGP_LONGJMP and VGP_SETJMP, since they're the same on all
Nicholas Nethercote [Mon, 16 May 2005 20:50:52 +0000 (20:50 +0000)] 
Remove VGP_LONGJMP and VGP_SETJMP, since they're the same on all
platforms.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3746

21 years agocore.h: remove some #include lines, move some others to more localised
Nicholas Nethercote [Mon, 16 May 2005 20:40:51 +0000 (20:40 +0000)] 
core.h: remove some #include lines, move some others to more localised
places.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3745

21 years agowibble
Julian Seward [Mon, 16 May 2005 18:30:40 +0000 (18:30 +0000)] 
wibble

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3744

21 years agoCFI reader: handle DW_CFA_GNU_window_save. Why is this sparc-specific
Julian Seward [Mon, 16 May 2005 18:25:09 +0000 (18:25 +0000)] 
CFI reader: handle DW_CFA_GNU_window_save.  Why is this sparc-specific
construct turning up in .so's in x86 SuSE 9.3 ?  Answers on a postcard
please.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3743

21 years agoMore FC4 jiggery-pokery.
Julian Seward [Mon, 16 May 2005 16:39:57 +0000 (16:39 +0000)] 
More FC4 jiggery-pokery.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3742

21 years agoDisable a bunch of suppressions made obsolete by r3740.
Julian Seward [Mon, 16 May 2005 16:39:28 +0000 (16:39 +0000)] 
Disable a bunch of suppressions made obsolete by r3740.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3741

21 years agoFix for sys_sysctl wrapper, from Ulrich Drepper.
Julian Seward [Mon, 16 May 2005 16:38:45 +0000 (16:38 +0000)] 
Fix for sys_sysctl wrapper, from Ulrich Drepper.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3740

21 years agoNo need to generate valgrind.h from valgrind.h.in. We can just use the
Nicholas Nethercote [Mon, 16 May 2005 13:31:23 +0000 (13:31 +0000)] 
No need to generate valgrind.h from valgrind.h.in.  We can just use the
compiler symbols rather than our own symbols for the architectures (we
already were, more or less).  This simplifies the build, hurrah!

I also inverted the sense of the NVALGRIND ifdefs, to make them easier to
read, and fixed up some comments.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3739

21 years agoCFI reader: partially handle DW_CFA_expression and
Julian Seward [Mon, 16 May 2005 11:48:34 +0000 (11:48 +0000)] 
CFI reader: partially handle DW_CFA_expression and
DW_CFA_def_cfa_expression in the sense that they are parsed correctly,
but the contained expression is ignored.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3738

21 years agoFix comment.
Julian Seward [Mon, 16 May 2005 11:47:17 +0000 (11:47 +0000)] 
Fix comment.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3737

21 years agoFill in the amd64 case for regaddr_from_tst.
Tom Hughes [Mon, 16 May 2005 07:31:57 +0000 (07:31 +0000)] 
Fill in the amd64 case for regaddr_from_tst.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3736

21 years agoUse movq to get the frame pointer on amd64.
Tom Hughes [Mon, 16 May 2005 07:31:05 +0000 (07:31 +0000)] 
Use movq to get the frame pointer on amd64.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3735

21 years agoComplete restructuring of include/ directory -- I forgot about the ARM
Nicholas Nethercote [Mon, 16 May 2005 03:36:19 +0000 (03:36 +0000)] 
Complete restructuring of include/ directory -- I forgot about the ARM
stuff.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3734

21 years agoRestructured the include/ directory.
Nicholas Nethercote [Mon, 16 May 2005 03:25:12 +0000 (03:25 +0000)] 
Restructured the include/ directory.

- All the subdirectories have gone: arm/, x86/, amd64/, linux/,
  x86-linux/, amd64-linux/, arm-linux/.

- The following files were moved out of those directories into include/:

    amd64-linux/vki_arch.h            --> vki-amd64-linux.h
    x86-linux/vki_arch.h              --> vki-x86-linux.h
    x86-linux/vki_arch_posixtypes.h   --> vki_posixtypes-x86-linux.h
    linux/vki.h                       --> vki-linux.h
    amd64-linux/vki_arch_posixtypes.h --> vki_posixtypes-amd64-linux.h

- The following very small files were inlined into tool.h using the "#if
  defined(VGP_x86)..." technique:

    x86/tool_arch.h
    arm/tool_arch.h
    amd64/tool_arch.h

  The same technique was used twice to include the appropriate
  vki-$PLATFORM and vki-$OS files into tool.h.

- The other files in those directories were removed.

- The build is much simpler, since we have 7(!) fewer Makefile.am files.
  Far fewer -I options are needed when compiling, too.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3733

21 years agoMoved the stack-walking macros into m_stacktrace.
Nicholas Nethercote [Mon, 16 May 2005 00:16:56 +0000 (00:16 +0000)] 
Moved the stack-walking macros into m_stacktrace.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3732

21 years agoMove some decls in vg_dwarf.c closer to the comment explaining them.
Nicholas Nethercote [Mon, 16 May 2005 00:04:02 +0000 (00:04 +0000)] 
Move some decls in vg_dwarf.c closer to the comment explaining them.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3731

21 years agoMoved VGA_(reg_addr_from_tst)() and VGA_R_{STACK,FRAME}_PTR into vg_symtab2.c.
Nicholas Nethercote [Mon, 16 May 2005 00:01:04 +0000 (00:01 +0000)] 
Moved VGA_(reg_addr_from_tst)() and VGA_R_{STACK,FRAME}_PTR into vg_symtab2.c.
Also made regaddr() and co. 64-bit clean (I hope I did that correctly).

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3730

21 years agoMove VGA_GET_REAL_{STACK,FRAME}_PTR into vg_symtab2.c.
Nicholas Nethercote [Sun, 15 May 2005 23:36:12 +0000 (23:36 +0000)] 
Move VGA_GET_REAL_{STACK,FRAME}_PTR into vg_symtab2.c.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3729

21 years agobugfix: don't print bogus "obj:" lines when generating suppressions.
Nicholas Nethercote [Sun, 15 May 2005 21:09:40 +0000 (21:09 +0000)] 
bugfix: don't print bogus "obj:" lines when generating suppressions.
Thanks to Julian for spotting and diagnosing it.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3728

21 years agoLet's assume the client's address space starts at 0 until we port
Nicholas Nethercote [Sun, 15 May 2005 21:03:42 +0000 (21:03 +0000)] 
Let's assume the client's address space starts at 0 until we port
to a platform for which that's not true.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3727

21 years agoImproved structure of LDT-related code:
Nicholas Nethercote [Sun, 15 May 2005 20:52:04 +0000 (20:52 +0000)] 
Improved structure of LDT-related code:
- one declarations from core.h removed, one moved to within m_syscalls.
- all the x86 LDT stuff made local to m_syscalls.  x86-linux/ldt.c removed
  as a result.  x86/state.c slimmed down, too.  x86/x86_private.h removed
  too.
- all the AMD64 LDT stuff was deleted, since it was all commented out.  It
  can be added back in later in the appropriate places if necessary.
  Thus amd64-linux/ldt.c and amd64/amd64_private.h were removed.
- other minor naming changes

I hope I didn't break AMD64 compilation.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3726

21 years agoCover the unknown-platform case.
Nicholas Nethercote [Sun, 15 May 2005 18:51:47 +0000 (18:51 +0000)] 
Cover the unknown-platform case.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3725

21 years agoRemove dead constant.
Nicholas Nethercote [Sun, 15 May 2005 17:38:56 +0000 (17:38 +0000)] 
Remove dead constant.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3724

21 years agoDon't mix backticks and apostrophes when quoting words -- eg. use 'foo'
Nicholas Nethercote [Sun, 15 May 2005 17:28:26 +0000 (17:28 +0000)] 
Don't mix backticks and apostrophes when quoting words -- eg. use 'foo'
rather than `foo', as www.cl.cam.ac.uk/~mgk25/ucs/quotes.html explains
we should (in more detail than you'd imagine was possible).  I did this
both in output messages and in some comments, for consistency.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3723

21 years agoglibc-2.3.5 fix (FC4)
Julian Seward [Sun, 15 May 2005 17:18:40 +0000 (17:18 +0000)] 
glibc-2.3.5 fix (FC4)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3722

21 years agoCFI reader: handle DW_CFA_def_cfa_offset_sf -- FC4 needs it
Julian Seward [Sun, 15 May 2005 17:14:28 +0000 (17:14 +0000)] 
CFI reader: handle DW_CFA_def_cfa_offset_sf -- FC4 needs it

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3721

21 years agoCleaned up the logging command-line option variables:
Nicholas Nethercote [Sun, 15 May 2005 16:22:58 +0000 (16:22 +0000)] 
Cleaned up the logging command-line option variables:
- renamed VG_(logging_to_filedes) as VG_(logging_to_socket), which is
  clearer
- no longer exporting clo_log_to, which avoids the confusion about whether
  it or VG_(logging_to_socket) actually controls where output goes
- couple of other minor cleanups

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3720

21 years agoAdd note to FAQ about unloaded shared objects and leak errors.
Nicholas Nethercote [Sun, 15 May 2005 14:49:24 +0000 (14:49 +0000)] 
Add note to FAQ about unloaded shared objects and leak errors.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3719

21 years agoRemove VGA_WORD_SIZE; we already have VG_WORDSIZE.
Nicholas Nethercote [Sun, 15 May 2005 03:56:26 +0000 (03:56 +0000)] 
Remove VGA_WORD_SIZE;  we already have VG_WORDSIZE.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3718

21 years agofixup headers a bit
Nicholas Nethercote [Sun, 15 May 2005 03:52:40 +0000 (03:52 +0000)] 
fixup headers a bit

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3717

21 years agoMove the symbol table reading closer to the debug info reading.
Nicholas Nethercote [Sun, 15 May 2005 03:38:53 +0000 (03:38 +0000)] 
Move the symbol table reading closer to the debug info reading.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3716

21 years agoBe more concise.
Nicholas Nethercote [Sun, 15 May 2005 03:30:56 +0000 (03:30 +0000)] 
Be more concise.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3715

21 years agoRemove call to VGA_(cleanup_thread)() from VG_(exit_thread)(), since
Nicholas Nethercote [Sat, 14 May 2005 23:28:42 +0000 (23:28 +0000)] 
Remove call to VGA_(cleanup_thread)() from VG_(exit_thread)(), since
mostly_clear_thread_record() calls it anyway.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3714

21 years agoRemove dead function.
Nicholas Nethercote [Sat, 14 May 2005 23:10:37 +0000 (23:10 +0000)] 
Remove dead function.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3713

21 years agoWhoops, a little premature in removing core_asm.h -- still need it
Nicholas Nethercote [Sat, 14 May 2005 21:51:49 +0000 (21:51 +0000)] 
Whoops, a little premature in removing core_asm.h -- still need it
because it #includes tool_asm.h.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3712

21 years agoCreated pub_core_transtab_asm.h, killing core_asm.h in the process.
Nicholas Nethercote [Sat, 14 May 2005 21:44:20 +0000 (21:44 +0000)] 
Created pub_core_transtab_asm.h, killing core_asm.h in the process.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3711

21 years agoModularised vg_hashtable.c as m_hashtable.
Nicholas Nethercote [Sat, 14 May 2005 21:28:43 +0000 (21:28 +0000)] 
Modularised vg_hashtable.c as m_hashtable.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3710

21 years agoRenamed the field Segment.symtab to Segment.seginfo to avoid confusion
Nicholas Nethercote [Sat, 14 May 2005 18:42:26 +0000 (18:42 +0000)] 
Renamed the field Segment.symtab to Segment.seginfo to avoid confusion
with SegInfo.symtab.

Also renamed VG_(symtab_{inc,dec}ref)() as VG_(seginfo_{inc,dec}ref)()
for the same reason.

Also renamed various SegInfo variables from "seg" to "si" to avoid
confusion with the many Segment variables called "seg".

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3709

21 years agoVG_(resolve_redir)() can now be made private to vg_redir.c.
Nicholas Nethercote [Sat, 14 May 2005 17:18:12 +0000 (17:18 +0000)] 
VG_(resolve_redir)() can now be made private to vg_redir.c.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3708

21 years agoMove resolve_redir_allsegs() to vg_redir.c, the only file that uses it,
Nicholas Nethercote [Sat, 14 May 2005 17:11:06 +0000 (17:11 +0000)] 
Move resolve_redir_allsegs() to vg_redir.c, the only file that uses it,
and make it private.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3707

21 years agoRemove a short sequence of code that was:
Nicholas Nethercote [Sat, 14 May 2005 17:00:25 +0000 (17:00 +0000)] 
Remove a short sequence of code that was:
(a) overly complex (the parens could be added by the sprintf);
(b) buggy (buf is used uninitialised if VG_(get_fnname) fails);
(c) redundant (the VG_(core_panic_at) prints the stack trace anyway).

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3706

21 years agoAdd test cases for MOVQ on amd64.
Julian Seward [Sat, 14 May 2005 11:18:31 +0000 (11:18 +0000)] 
Add test cases for MOVQ on amd64.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3705

21 years agoFix behaviour of MOVQ on amd64.
Julian Seward [Sat, 14 May 2005 11:17:25 +0000 (11:17 +0000)] 
Fix behaviour of MOVQ on amd64.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1198

21 years agoNo wonder I couldn't figure out exactly how movq was supposed to
Julian Seward [Sat, 14 May 2005 02:08:48 +0000 (02:08 +0000)] 
No wonder I couldn't figure out exactly how movq was supposed to
behave ..  no test cases!  Now fixed.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3704

21 years agoFinally fix the behaviour of MOVQ (xmm -> xmm).
Julian Seward [Sat, 14 May 2005 02:04:12 +0000 (02:04 +0000)] 
Finally fix the behaviour of MOVQ (xmm -> xmm).

git-svn-id: svn://svn.valgrind.org/vex/trunk@1197

21 years agoHandle NegF64.
Julian Seward [Sat, 14 May 2005 02:02:50 +0000 (02:02 +0000)] 
Handle NegF64.

git-svn-id: svn://svn.valgrind.org/vex/trunk@1196

21 years agoModularised vg_transtab.c as m_transtab.
Nicholas Nethercote [Fri, 13 May 2005 23:40:55 +0000 (23:40 +0000)] 
Modularised vg_transtab.c as m_transtab.

Renamed add_to_trans_tab() as add_to_transtab() for consistency
with the other names.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3703

21 years agoDon't crash when dealing with dealing with naughty programs that set
Julian Seward [Fri, 13 May 2005 23:32:43 +0000 (23:32 +0000)] 
Don't crash when dealing with dealing with naughty programs that set
argv[] elements to NULL.  Bad bad GHC ...

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3702

21 years agoForgot to add pub_core_translate.h to the Makefile.am.
Nicholas Nethercote [Fri, 13 May 2005 23:14:40 +0000 (23:14 +0000)] 
Forgot to add pub_core_translate.h to the Makefile.am.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3701

21 years agoModularised vg_translate.c as m_translate.
Nicholas Nethercote [Fri, 13 May 2005 23:11:40 +0000 (23:11 +0000)] 
Modularised vg_translate.c as m_translate.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3700

21 years agoImport vki_unistd.h on a need-to-know basis, rather than
Nicholas Nethercote [Fri, 13 May 2005 22:18:47 +0000 (22:18 +0000)] 
Import vki_unistd.h on a need-to-know basis, rather than
universally through core.h.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3699

21 years agoVGA_(terminate) doesn't need to be exported.
Nicholas Nethercote [Fri, 13 May 2005 22:09:26 +0000 (22:09 +0000)] 
VGA_(terminate) doesn't need to be exported.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3698

21 years agoVGA_(restart_syscall)() doesn't need to be exported.
Nicholas Nethercote [Fri, 13 May 2005 22:07:12 +0000 (22:07 +0000)] 
VGA_(restart_syscall)() doesn't need to be exported.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3697

21 years agoUse HChar for format strings to printf-like functions.
Nicholas Nethercote [Fri, 13 May 2005 21:49:30 +0000 (21:49 +0000)] 
Use HChar for format strings to printf-like functions.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3696

21 years agoRemove all traces of the now-defunct ThreadArchAux type.
Nicholas Nethercote [Fri, 13 May 2005 21:44:07 +0000 (21:44 +0000)] 
Remove all traces of the now-defunct ThreadArchAux type.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3695

21 years agoRemove unnecessary #includes.
Nicholas Nethercote [Fri, 13 May 2005 21:42:08 +0000 (21:42 +0000)] 
Remove unnecessary #includes.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@3694