* [RFA] gprof static call support for MIPS
@ 2002-01-20 15:54 Jason R Thorpe
2002-01-20 15:59 ` Daniel Jacobowitz
2002-01-26 2:28 ` Nick Clifton
0 siblings, 2 replies; 7+ messages in thread
From: Jason R Thorpe @ 2002-01-20 15:54 UTC (permalink / raw)
To: binutils
[-- Attachment #1: Type: text/plain, Size: 422 bytes --]
The following adds support for the gprof -c option (static call
graph by hueristic) on the MIPS. It was derived from similar code
for the Alpha.
OK to commit?
* mips.c: New file.
* Makefile.am (sources): Add mips.c.
(mips.o): New rule.
* Makefile.in: Regenerate.
* corefile.c: Update copyright years.
(find_call): Call mips_find_call for bfd_arch_mips.
--
-- Jason R. Thorpe <thorpej@wasabisystems.com>
[-- Attachment #2: gprof-mips-patch --]
[-- Type: text/plain, Size: 9333 bytes --]
Index: Makefile.am
===================================================================
RCS file: /cvs/src/src/gprof/Makefile.am,v
retrieving revision 1.11
diff -c -r1.11 Makefile.am
*** Makefile.am 2001/10/02 06:03:15 1.11
--- Makefile.am 2002/01/20 23:04:07
***************
*** 23,29 ****
sources = basic_blocks.c call_graph.c cg_arcs.c cg_dfn.c \
cg_print.c corefile.c gmon_io.c gprof.c hertz.c hist.c source.c \
search_list.c symtab.c sym_ids.c utils.c \
! i386.c alpha.c vax.c tahoe.c sparc.c
gprof_SOURCES = $(sources) flat_bl.c bsd_callg_bl.c fsf_callg_bl.c
gprof_DEPENDENCIES = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLDEPS)
gprof_LDADD = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLLIBS)
--- 23,29 ----
sources = basic_blocks.c call_graph.c cg_arcs.c cg_dfn.c \
cg_print.c corefile.c gmon_io.c gprof.c hertz.c hist.c source.c \
search_list.c symtab.c sym_ids.c utils.c \
! i386.c alpha.c vax.c tahoe.c sparc.c mips.c
gprof_SOURCES = $(sources) flat_bl.c bsd_callg_bl.c fsf_callg_bl.c
gprof_DEPENDENCIES = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLDEPS)
gprof_LDADD = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLLIBS)
***************
*** 216,221 ****
--- 216,226 ----
$(INCDIR)/symcat.h source.h search_list.h corefile.h \
hist.h
sparc.o: sparc.c gprof.h $(INCDIR)/ansidecl.h $(BFDDIR)/sysdep.h \
+ ../bfd/config.h $(INCDIR)/fopen-same.h $(INCDIR)/filenames.h \
+ gconfig.h $(INCDIR)/bin-bugs.h cg_arcs.h symtab.h ../bfd/bfd.h \
+ $(INCDIR)/symcat.h source.h search_list.h corefile.h \
+ hist.h
+ mips.o: mips.c gprof.h $(INCDIR)/ansidecl.h $(BFDDIR)/sysdep.h \
../bfd/config.h $(INCDIR)/fopen-same.h $(INCDIR)/filenames.h \
gconfig.h $(INCDIR)/bin-bugs.h cg_arcs.h symtab.h ../bfd/bfd.h \
$(INCDIR)/symcat.h source.h search_list.h corefile.h \
Index: Makefile.in
===================================================================
RCS file: /cvs/src/src/gprof/Makefile.in,v
retrieving revision 1.16
diff -c -r1.16 Makefile.in
*** Makefile.in 2001/10/02 06:03:15 1.16
--- Makefile.in 2002/01/20 23:04:07
***************
*** 123,129 ****
sources = basic_blocks.c call_graph.c cg_arcs.c cg_dfn.c \
cg_print.c corefile.c gmon_io.c gprof.c hertz.c hist.c source.c \
search_list.c symtab.c sym_ids.c utils.c \
! i386.c alpha.c vax.c tahoe.c sparc.c
gprof_SOURCES = $(sources) flat_bl.c bsd_callg_bl.c fsf_callg_bl.c
gprof_DEPENDENCIES = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLDEPS)
--- 123,129 ----
sources = basic_blocks.c call_graph.c cg_arcs.c cg_dfn.c \
cg_print.c corefile.c gmon_io.c gprof.c hertz.c hist.c source.c \
search_list.c symtab.c sym_ids.c utils.c \
! i386.c alpha.c vax.c tahoe.c sparc.c mips.c
gprof_SOURCES = $(sources) flat_bl.c bsd_callg_bl.c fsf_callg_bl.c
gprof_DEPENDENCIES = ../bfd/libbfd.la ../libiberty/libiberty.a $(INTLDEPS)
***************
*** 167,174 ****
corefile.$(OBJEXT) gmon_io.$(OBJEXT) gprof.$(OBJEXT) hertz.$(OBJEXT) \
hist.$(OBJEXT) source.$(OBJEXT) search_list.$(OBJEXT) symtab.$(OBJEXT) \
sym_ids.$(OBJEXT) utils.$(OBJEXT) i386.$(OBJEXT) alpha.$(OBJEXT) \
! vax.$(OBJEXT) tahoe.$(OBJEXT) sparc.$(OBJEXT) flat_bl.$(OBJEXT) \
! bsd_callg_bl.$(OBJEXT) fsf_callg_bl.$(OBJEXT)
gprof_LDFLAGS =
CFLAGS = @CFLAGS@
COMPILE = $(CC) $(DEFS) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS)
--- 167,174 ----
corefile.$(OBJEXT) gmon_io.$(OBJEXT) gprof.$(OBJEXT) hertz.$(OBJEXT) \
hist.$(OBJEXT) source.$(OBJEXT) search_list.$(OBJEXT) symtab.$(OBJEXT) \
sym_ids.$(OBJEXT) utils.$(OBJEXT) i386.$(OBJEXT) alpha.$(OBJEXT) \
! vax.$(OBJEXT) tahoe.$(OBJEXT) sparc.$(OBJEXT) mips.$(OBJEXT) \
! flat_bl.$(OBJEXT) bsd_callg_bl.$(OBJEXT) fsf_callg_bl.$(OBJEXT)
gprof_LDFLAGS =
CFLAGS = @CFLAGS@
COMPILE = $(CC) $(DEFS) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS)
***************
*** 879,884 ****
--- 879,889 ----
$(INCDIR)/symcat.h source.h search_list.h corefile.h \
hist.h
sparc.o: sparc.c gprof.h $(INCDIR)/ansidecl.h $(BFDDIR)/sysdep.h \
+ ../bfd/config.h $(INCDIR)/fopen-same.h $(INCDIR)/filenames.h \
+ gconfig.h $(INCDIR)/bin-bugs.h cg_arcs.h symtab.h ../bfd/bfd.h \
+ $(INCDIR)/symcat.h source.h search_list.h corefile.h \
+ hist.h
+ mips.o: mips.c gprof.h $(INCDIR)/ansidecl.h $(BFDDIR)/sysdep.h \
../bfd/config.h $(INCDIR)/fopen-same.h $(INCDIR)/filenames.h \
gconfig.h $(INCDIR)/bin-bugs.h cg_arcs.h symtab.h ../bfd/bfd.h \
$(INCDIR)/symcat.h source.h search_list.h corefile.h \
Index: corefile.c
===================================================================
RCS file: /cvs/src/src/gprof/corefile.c,v
retrieving revision 1.7
diff -c -r1.7 corefile.c
*** corefile.c 2001/03/14 03:14:56 1.7
--- corefile.c 2002/01/20 23:04:08
***************
*** 1,6 ****
/* corefile.c
! Copyright 2000, 2001 Free Software Foundation, Inc.
This file is part of GNU Binutils.
--- 1,6 ----
/* corefile.c
! Copyright 2000, 2001, 2002 Free Software Foundation, Inc.
This file is part of GNU Binutils.
***************
*** 48,53 ****
--- 48,54 ----
extern void vax_find_call PARAMS ((Sym *, bfd_vma, bfd_vma));
extern void tahoe_find_call PARAMS ((Sym *, bfd_vma, bfd_vma));
extern void sparc_find_call PARAMS ((Sym *, bfd_vma, bfd_vma));
+ extern void mips_find_call PARAMS ((Sym *, bfd_vma, bfd_vma));
static void
DEFUN (read_function_mappings, (filename), const char *filename)
***************
*** 259,264 ****
--- 260,269 ----
case bfd_arch_tahoe:
tahoe_find_call (parent, p_lowpc, p_highpc);
+ break;
+
+ case bfd_arch_mips:
+ mips_find_call (parent, p_lowpc, p_highpc);
break;
default:
Index: mips.c
===================================================================
RCS file: mips.c
diff -N mips.c
*** /dev/null Tue May 5 13:32:27 1998
--- mips.c Sun Jan 20 15:04:08 2002
***************
*** 0 ****
--- 1,102 ----
+ /*
+ * Copyright (c) 1983, 1998 Regents of the University of California.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms are permitted
+ * provided that: (1) source distributions retain this entire copyright
+ * notice and comment, and (2) distributions including binaries display
+ * the following acknowledgement: ``This product includes software
+ * developed by the University of California, Berkeley and its contributors''
+ * in the documentation or other materials provided with the distribution
+ * and in all advertising materials mentioning features or use of this
+ * software. Neither the name of the University nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
+ * IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
+ */
+ #include "gprof.h"
+ #include "cg_arcs.h"
+ #include "corefile.h"
+ #include "hist.h"
+ #include "symtab.h"
+
+ static Sym indirect_child;
+
+ void mips_find_call PARAMS ((Sym *, bfd_vma, bfd_vma));
+
+ void
+ mips_find_call (parent, p_lowpc, p_highpc)
+ Sym *parent;
+ bfd_vma p_lowpc;
+ bfd_vma p_highpc;
+ {
+ bfd_vma pc, dest_pc;
+ unsigned int op;
+ int offset;
+ Sym *child;
+ static bool inited = FALSE;
+
+ if (!inited)
+ {
+ inited = TRUE;
+ sym_init (&indirect_child);
+ indirect_child.name = _("<indirect child>");
+ indirect_child.cg.prop.fract = 1.0;
+ indirect_child.cg.cyc.head = &indirect_child;
+ }
+
+ if (!core_text_space)
+ {
+ return;
+ }
+ if (p_lowpc < s_lowpc)
+ {
+ p_lowpc = s_lowpc;
+ }
+ if (p_highpc > s_highpc)
+ {
+ p_highpc = s_highpc;
+ }
+ DBG (CALLDEBUG, printf (_("[find_call] %s: 0x%lx to 0x%lx\n"),
+ parent->name, (unsigned long) p_lowpc,
+ (unsigned long) p_highpc));
+ for (pc = p_lowpc; pc < p_highpc; pc += 4)
+ {
+ op = bfd_get_32 (core_bfd, &((char *)core_text_space)[pc - s_lowpc]);
+ if ((op & 0xfc000000) == 0x0c000000)
+ {
+ /* This is a "jal" instruction. Check that the destination
+ is the address of a function. */
+ DBG (CALLDEBUG,
+ printf (_("[find_call] 0x%lx: jal"), (unsigned long) pc));
+ offset = (op & 0x03ffffff) << 2;
+ dest_pc = (pc & ~0xfffffff) | offset;
+ if (dest_pc >= s_lowpc && dest_pc <= s_highpc)
+ {
+ child = sym_lookup (&symtab, dest_pc);
+ DBG (CALLDEBUG,
+ printf (" 0x%lx\t; name=%s, addr=0x%lx",
+ (unsigned long) dest_pc, child->name,
+ (unsigned long) child->addr));
+ if (child->addr == dest_pc)
+ {
+ DBG (CALLDEBUG, printf ("\n"));
+ /* a hit: */
+ arc_add (parent, child, (unsigned long) 0);
+ continue;
+ }
+ }
+ /* Something funny going on. */
+ DBG (CALLDEBUG, printf ("\tbut it's a botch\n"));
+ }
+ else if ((op & 0xfc00f83f) == 0x0000f809)
+ {
+ /* This is a "jalr" instruction (indirect call). */
+ DBG (CALLDEBUG,
+ printf (_("[find_call] 0x%lx: jalr\n"), (unsigned long) pc));
+ arc_add (parent, &indirect_child, (unsigned long) 0);
+ }
+ }
+ }
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-20 15:54 [RFA] gprof static call support for MIPS Jason R Thorpe
@ 2002-01-20 15:59 ` Daniel Jacobowitz
2002-01-20 16:12 ` Jason R Thorpe
2002-01-26 2:28 ` Nick Clifton
1 sibling, 1 reply; 7+ messages in thread
From: Daniel Jacobowitz @ 2002-01-20 15:59 UTC (permalink / raw)
To: binutils; +Cc: Jason R Thorpe
On Sun, Jan 20, 2002 at 03:08:22PM -0800, Jason R Thorpe wrote:
> The following adds support for the gprof -c option (static call
> graph by hueristic) on the MIPS. It was derived from similar code
> for the Alpha.
>
> OK to commit?
>
> * mips.c: New file.
> * Makefile.am (sources): Add mips.c.
> (mips.o): New rule.
> * Makefile.in: Regenerate.
> * corefile.c: Update copyright years.
> (find_call): Call mips_find_call for bfd_arch_mips.
>
> --
> -- Jason R. Thorpe <thorpej@wasabisystems.com>
> + for (pc = p_lowpc; pc < p_highpc; pc += 4)
> + {
> + op = bfd_get_32 (core_bfd, &((char *)core_text_space)[pc - s_lowpc]);
> + if ((op & 0xfc000000) == 0x0c000000)
> + {
> + /* This is a "jal" instruction. Check that the destination
> + is the address of a function. */
> + DBG (CALLDEBUG,
> + printf (_("[find_call] 0x%lx: jal"), (unsigned long) pc));
> + offset = (op & 0x03ffffff) << 2;
> + dest_pc = (pc & ~0xfffffff) | offset;
> + if (dest_pc >= s_lowpc && dest_pc <= s_highpc)
> + {
> + child = sym_lookup (&symtab, dest_pc);
> + DBG (CALLDEBUG,
> + printf (" 0x%lx\t; name=%s, addr=0x%lx",
> + (unsigned long) dest_pc, child->name,
> + (unsigned long) child->addr));
> + if (child->addr == dest_pc)
> + {
> + DBG (CALLDEBUG, printf ("\n"));
> + /* a hit: */
> + arc_add (parent, child, (unsigned long) 0);
> + continue;
> + }
> + }
> + /* Something funny going on. */
> + DBG (CALLDEBUG, printf ("\tbut it's a botch\n"));
> + }
> + else if ((op & 0xfc00f83f) == 0x0000f809)
> + {
> + /* This is a "jalr" instruction (indirect call). */
> + DBG (CALLDEBUG,
> + printf (_("[find_call] 0x%lx: jalr\n"), (unsigned long) pc));
> + arc_add (parent, &indirect_child, (unsigned long) 0);
> + }
> + }
> + }
Is this going to do any good for NetBSD? I thought I remembered you
saying that NetBSD/MIPS used -mabicalls, when you posted the profiling
patch. In that case all interesting calls are indirect!
Then again I don't see abicalls set for netbsd in the community GCC
tree.
--
Daniel Jacobowitz Carnegie Mellon University
MontaVista Software Debian GNU/Linux Developer
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-20 15:59 ` Daniel Jacobowitz
@ 2002-01-20 16:12 ` Jason R Thorpe
2002-01-20 16:22 ` Daniel Jacobowitz
0 siblings, 1 reply; 7+ messages in thread
From: Jason R Thorpe @ 2002-01-20 16:12 UTC (permalink / raw)
To: binutils
On Sun, Jan 20, 2002 at 06:26:05PM -0500, Daniel Jacobowitz wrote:
> Is this going to do any good for NetBSD? I thought I remembered you
> saying that NetBSD/MIPS used -mabicalls, when you posted the profiling
> patch. In that case all interesting calls are indirect!
The kernel does not use abicalls. We use gprof on kernels, too.
> Then again I don't see abicalls set for netbsd in the community GCC
> tree.
I have a new NetBSD/mips configuration for GCC that is approved, but
I'm waiting for the other bits of my patch to be approved so I can
commit it (the MIPS portion depends on changes to the shared NetBSD ELF
configuration bits). But, in any case, yes, abicalls is the default
for userland code.
--
-- Jason R. Thorpe <thorpej@wasabisystems.com>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-20 16:12 ` Jason R Thorpe
@ 2002-01-20 16:22 ` Daniel Jacobowitz
2002-01-25 9:03 ` Jason R Thorpe
0 siblings, 1 reply; 7+ messages in thread
From: Daniel Jacobowitz @ 2002-01-20 16:22 UTC (permalink / raw)
To: Jason R Thorpe, binutils
On Sun, Jan 20, 2002 at 03:54:41PM -0800, Jason R Thorpe wrote:
> On Sun, Jan 20, 2002 at 06:26:05PM -0500, Daniel Jacobowitz wrote:
>
> > Is this going to do any good for NetBSD? I thought I remembered you
> > saying that NetBSD/MIPS used -mabicalls, when you posted the profiling
> > patch. In that case all interesting calls are indirect!
>
> The kernel does not use abicalls. We use gprof on kernels, too.
Ah, yes. Ditto for the Linux kernel.
> > Then again I don't see abicalls set for netbsd in the community GCC
> > tree.
>
> I have a new NetBSD/mips configuration for GCC that is approved, but
> I'm waiting for the other bits of my patch to be approved so I can
> commit it (the MIPS portion depends on changes to the shared NetBSD ELF
> configuration bits). But, in any case, yes, abicalls is the default
> for userland code.
OK, that makes sense. I don't know if it's possible to reconstitute
some of the GOT-based calls in the callgraph, but it may be... I'll
have to think about it.
--
Daniel Jacobowitz Carnegie Mellon University
MontaVista Software Debian GNU/Linux Developer
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-20 16:22 ` Daniel Jacobowitz
@ 2002-01-25 9:03 ` Jason R Thorpe
2002-01-25 11:59 ` Daniel Jacobowitz
0 siblings, 1 reply; 7+ messages in thread
From: Jason R Thorpe @ 2002-01-25 9:03 UTC (permalink / raw)
To: binutils
On Sun, Jan 20, 2002 at 06:59:32PM -0500, Daniel Jacobowitz wrote:
> On Sun, Jan 20, 2002 at 03:54:41PM -0800, Jason R Thorpe wrote:
> > On Sun, Jan 20, 2002 at 06:26:05PM -0500, Daniel Jacobowitz wrote:
> >
> > > Is this going to do any good for NetBSD? I thought I remembered you
> > > saying that NetBSD/MIPS used -mabicalls, when you posted the profiling
> > > patch. In that case all interesting calls are indirect!
> >
> > The kernel does not use abicalls. We use gprof on kernels, too.
>
> Ah, yes. Ditto for the Linux kernel.
So, have we come to the conclusion that, while not perfect, it does
provide a useful feature for certain applications?
And, if so, can I have formal approval for the patch so that I may
commit it? :-)
--
-- Jason R. Thorpe <thorpej@wasabisystems.com>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-25 9:03 ` Jason R Thorpe
@ 2002-01-25 11:59 ` Daniel Jacobowitz
0 siblings, 0 replies; 7+ messages in thread
From: Daniel Jacobowitz @ 2002-01-25 11:59 UTC (permalink / raw)
To: binutils; +Cc: Jason R Thorpe
On Fri, Jan 25, 2002 at 08:46:16AM -0800, Jason R Thorpe wrote:
> On Sun, Jan 20, 2002 at 06:59:32PM -0500, Daniel Jacobowitz wrote:
>
> > On Sun, Jan 20, 2002 at 03:54:41PM -0800, Jason R Thorpe wrote:
> > > On Sun, Jan 20, 2002 at 06:26:05PM -0500, Daniel Jacobowitz wrote:
> > >
> > > > Is this going to do any good for NetBSD? I thought I remembered you
> > > > saying that NetBSD/MIPS used -mabicalls, when you posted the profiling
> > > > patch. In that case all interesting calls are indirect!
> > >
> > > The kernel does not use abicalls. We use gprof on kernels, too.
> >
> > Ah, yes. Ditto for the Linux kernel.
>
> So, have we come to the conclusion that, while not perfect, it does
> provide a useful feature for certain applications?
>
> And, if so, can I have formal approval for the patch so that I may
> commit it? :-)
To answer your questions, yes, and ask someone else :)
--
Daniel Jacobowitz Carnegie Mellon University
MontaVista Software Debian GNU/Linux Developer
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [RFA] gprof static call support for MIPS
2002-01-20 15:54 [RFA] gprof static call support for MIPS Jason R Thorpe
2002-01-20 15:59 ` Daniel Jacobowitz
@ 2002-01-26 2:28 ` Nick Clifton
1 sibling, 0 replies; 7+ messages in thread
From: Nick Clifton @ 2002-01-26 2:28 UTC (permalink / raw)
To: thorpej; +Cc: binutils
Hi Jason,
> * mips.c: New file.
> * Makefile.am (sources): Add mips.c.
> (mips.o): New rule.
> * Makefile.in: Regenerate.
> * corefile.c: Update copyright years.
> (find_call): Call mips_find_call for bfd_arch_mips.
Approved - please apply
Cheers
Nick
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2002-01-26 10:25 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-01-20 15:54 [RFA] gprof static call support for MIPS Jason R Thorpe
2002-01-20 15:59 ` Daniel Jacobowitz
2002-01-20 16:12 ` Jason R Thorpe
2002-01-20 16:22 ` Daniel Jacobowitz
2002-01-25 9:03 ` Jason R Thorpe
2002-01-25 11:59 ` Daniel Jacobowitz
2002-01-26 2:28 ` Nick Clifton
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).