* [RFA] implement gcore on hp/ux
@ 2006-01-10 17:18 Joel Brobecker
2006-01-10 20:29 ` Mark Kettenis
0 siblings, 1 reply; 6+ messages in thread
From: Joel Brobecker @ 2006-01-10 17:18 UTC (permalink / raw)
To: gdb-patches
[-- Attachment #1: Type: text/plain, Size: 3319 bytes --]
Hello,
The attached patch aims at implementing the gcore command on hp/ux.
Contrary to the current gcore implementation where we write the core
file ourselves, the implementation on hp/ux relies on the TT_PROC_CORE
ttrace command.
In order to make the gcore command a bit more generic, I added a new
method in the target vector which, when set, is responsible for writing
the core file.
The core of gcore.c that used to write the core file has been moved
to gcore-elf.c, and the exec_ops target vector is updated there.
Since there is now a second method to update (the one we just added),
I added a new function in exec.c with a corresponding declaration in
defs.h (besides the one we already have).
gcore.o is now always part of the build. When the target does not
support core files, the debugger now prints that this command is
not supported on this platform.
All builds that were including gcore.o are now including gcore-elf.o.
Including this file causes us to update the exec target ops as we used
to, and hence hooks up the "elf" implementation.
For the ttrace target, it's straighforward...
2005-01-10 Joel Brobecker <brobecker@adacore.com>
* target.h (target_ops): New method to_write_core_file.
(target_write_core_file_p): New macro.
(target_write_core_file): New macro.
* target.c (update_current_target): Add heritance of to_write_core_file.
* exec.c (exec_set_write_core_file): New function.
* defs.h (exec_set_write_core_file): Add declaration.
* gcore.c: Extract out the code writing the core file and move it
to gcore-elf.c.
* gcore-elf.c: New file, mostly extracted from gcore.o.
* Makefile.in (COMMON_OBS): Add gcore.o.
(gcore.o): Update dependencies.
(gcore-elf.o): Add rule.
* inf-ttrace (inf_ttrace_write_core_file): New function.
(inf_ttrace_target): set the to_write_core_file method.
* config/alpha/fbsd.mh (NATDEPFILES): Likewise.
* config/arm/linux.mh (NATDEPFILES): Likewise.
* config/i386/fbsd.mh (NATDEPFILES): Likewise.
* config/i386/fbsd64.mh (NATDEPFILES): Likewise.
* config/i386/i386sol2.mh (NATDEPFILES): Likewise.
* config/i386/i386v42mp.mh (NATDEPFILES): Likewise.
* config/i386/linux.mh (NATDEPFILES): Likewise.
* config/i386/linux64.mh (NATDEPFILES): Likewise.
* config/i386/sol2-64.mh (NATDEPFILES): Likewise.
* config/ia64/linux.mh (NATDEPFILES): Likewise.
* config/m32r/linux.mh (NATDEPFILES): Likewise.
* config/m68k/linux.mh (NATDEPFILES): Likewise.
* config/mips/linux.mh (NATDEPFILES): Likewise.
* config/pa/linux.mh (NATDEPFILES): Likewise.
* config/powerpc/linux.mh (NATDEPFILES): Likewise.
* config/powerpc/ppc64-linux.mh (NATDEPFILES): Likewise.
* config/s390/s390.mh (NATDEPFILES): Likewise.
* config/sparc/fbsd.mh (NATDEPFILES): Likewise.
* config/sparc/linux.mh (NATDEPFILES): Likewise.
* config/sparc/linux64.mh (NATDEPFILES): Likewise.
* config/sparc/sol2.mh (NATDEPFILES): Likewise.
Tested on x86-linux and pa32-hpux. No regression on the linux
side. On the hp/ux side, we get improvments from auxv.exp and
gcore.exp now passes.
Comments? Ok to apply?
Thanks,
--
Joel
[-- Attachment #2: gcore-hpux.diff --]
[-- Type: text/plain, Size: 33306 bytes --]
Index: target.h
===================================================================
RCS file: /cvs/src/src/gdb/target.h,v
retrieving revision 1.77
diff -u -p -r1.77 target.h
--- target.h 17 Dec 2005 22:34:03 -0000 1.77
+++ target.h 10 Jan 2006 16:03:28 -0000
@@ -398,6 +398,7 @@ struct target_ops
void (*to_async) (void (*cb) (enum inferior_event_type, void *context),
void *context);
int to_async_mask_value;
+ void (*to_write_core_file) (int pid, const char *filename);
int (*to_find_memory_regions) (int (*) (CORE_ADDR,
unsigned long,
int, int, int,
@@ -968,6 +969,13 @@ extern void (*deprecated_target_new_objf
#define target_pid_to_exec_file(pid) \
(current_target.to_pid_to_exec_file) (pid)
+/* Dump the inferior memory in a core file. */
+
+#define target_write_core_file_p \
+ (current_target.to_write_core_file != NULL)
+
+#define target_write_core_file(pid, filename) \
+ (current_target.to_write_core_file (pid, filename))
/*
* Iterator function for target memory regions.
* Calls a callback function once for each memory region 'mapped'
Index: target.c
===================================================================
RCS file: /cvs/src/src/gdb/target.c,v
retrieving revision 1.113
diff -u -p -r1.113 target.c
--- target.c 17 Dec 2005 22:34:03 -0000 1.113
+++ target.c 10 Jan 2006 16:03:27 -0000
@@ -453,6 +453,7 @@ update_current_target (void)
INHERIT (to_is_async_p, t);
INHERIT (to_async, t);
INHERIT (to_async_mask_value, t);
+ INHERIT (to_write_core_file, t);
INHERIT (to_find_memory_regions, t);
INHERIT (to_make_corefile_notes, t);
INHERIT (to_get_thread_local_address, t);
Index: exec.c
===================================================================
RCS file: /cvs/src/src/gdb/exec.c,v
retrieving revision 1.59
diff -u -p -r1.59 exec.c
--- exec.c 17 Dec 2005 22:33:59 -0000 1.59
+++ exec.c 10 Jan 2006 16:02:27 -0000
@@ -699,6 +699,14 @@ ignore (CORE_ADDR addr, bfd_byte *conten
return 0;
}
+/* Set the write_core_file method in the exec target vector. */
+
+void
+exec_set_write_core_file (void (*func) (int, const char *))
+{
+ exec_ops.to_write_core_file = func;
+}
+
/* Find mapped memory. */
extern void
Index: defs.h
===================================================================
RCS file: /cvs/src/src/gdb/defs.h,v
retrieving revision 1.190
diff -u -p -r1.190 defs.h
--- defs.h 17 Dec 2005 22:33:59 -0000 1.190
+++ defs.h 10 Jan 2006 16:02:19 -0000
@@ -623,6 +623,8 @@ extern void exec_set_section_offsets (bf
bfd_signed_vma data_off,
bfd_signed_vma bss_off);
+extern void exec_set_write_core_file (void (*func) (int, const char *));
+
/* Take over the 'find_mapped_memory' vector from exec.c. */
extern void exec_set_find_memory_regions (int (*) (int (*) (CORE_ADDR,
unsigned long,
Index: gcore.c
===================================================================
RCS file: /cvs/src/src/gdb/gcore.c,v
retrieving revision 1.18
diff -u -p -r1.18 gcore.c
--- gcore.c 17 Dec 2005 22:33:59 -0000 1.18
+++ gcore.c 10 Jan 2006 16:02:30 -0000
@@ -20,38 +20,24 @@
Boston, MA 02110-1301, USA. */
#include "defs.h"
-#include "elf-bfd.h"
-#include "infcall.h"
#include "inferior.h"
-#include "gdbcore.h"
-#include "objfiles.h"
-#include "symfile.h"
-
-#include "cli/cli-decode.h"
-
-#include "gdb_assert.h"
-
-static char *default_gcore_target (void);
-static enum bfd_architecture default_gcore_arch (void);
-static unsigned long default_gcore_mach (void);
-static int gcore_memory_sections (bfd *);
+#include "target.h"
+#include "command.h"
/* Generate a core file from the inferior process. */
static void
gcore_command (char *args, int from_tty)
{
- struct cleanup *old_chain;
char *corefilename, corefilename_buffer[40];
- asection *note_sec = NULL;
- bfd *obfd;
- void *note_data = NULL;
- int note_size = 0;
/* No use generating a corefile without a target process. */
if (!target_has_execution)
noprocess ();
+ if (!target_write_core_file_p)
+ error ("Command not supported on this platform");
+
if (args && *args)
corefilename = args;
else
@@ -61,431 +47,7 @@ gcore_command (char *args, int from_tty)
corefilename = corefilename_buffer;
}
- if (info_verbose)
- fprintf_filtered (gdb_stdout,
- "Opening corefile '%s' for output.\n", corefilename);
-
- /* Open the output file. */
- obfd = bfd_openw (corefilename, default_gcore_target ());
- if (!obfd)
- error (_("Failed to open '%s' for output."), corefilename);
-
- /* Need a cleanup that will close the file (FIXME: delete it?). */
- old_chain = make_cleanup_bfd_close (obfd);
-
- bfd_set_format (obfd, bfd_core);
- bfd_set_arch_mach (obfd, default_gcore_arch (), default_gcore_mach ());
-
- /* An external target method must build the notes section. */
- note_data = target_make_corefile_notes (obfd, ¬e_size);
-
- /* Create the note section. */
- if (note_data != NULL && note_size != 0)
- {
- note_sec = bfd_make_section_anyway (obfd, "note0");
- if (note_sec == NULL)
- error (_("Failed to create 'note' section for corefile: %s"),
- bfd_errmsg (bfd_get_error ()));
-
- bfd_set_section_vma (obfd, note_sec, 0);
- bfd_set_section_flags (obfd, note_sec,
- SEC_HAS_CONTENTS | SEC_READONLY | SEC_ALLOC);
- bfd_set_section_alignment (obfd, note_sec, 0);
- bfd_set_section_size (obfd, note_sec, note_size);
- }
-
- /* Now create the memory/load sections. */
- if (gcore_memory_sections (obfd) == 0)
- error (_("gcore: failed to get corefile memory sections from target."));
-
- /* Write out the contents of the note section. */
- if (note_data != NULL && note_size != 0)
- {
- if (!bfd_set_section_contents (obfd, note_sec, note_data, 0, note_size))
- warning (_("writing note section (%s)"), bfd_errmsg (bfd_get_error ()));
- }
-
- /* Succeeded. */
- fprintf_filtered (gdb_stdout, "Saved corefile %s\n", corefilename);
-
- /* Clean-ups will close the output file and free malloc memory. */
- do_cleanups (old_chain);
- return;
-}
-
-static unsigned long
-default_gcore_mach (void)
-{
-#if 1 /* See if this even matters... */
- return 0;
-#else
-#ifdef TARGET_ARCHITECTURE
- const struct bfd_arch_info *bfdarch = TARGET_ARCHITECTURE;
-
- if (bfdarch != NULL)
- return bfdarch->mach;
-#endif /* TARGET_ARCHITECTURE */
- if (exec_bfd == NULL)
- error (_("Can't find default bfd machine type (need execfile)."));
-
- return bfd_get_mach (exec_bfd);
-#endif /* 1 */
-}
-
-static enum bfd_architecture
-default_gcore_arch (void)
-{
-#ifdef TARGET_ARCHITECTURE
- const struct bfd_arch_info * bfdarch = TARGET_ARCHITECTURE;
-
- if (bfdarch != NULL)
- return bfdarch->arch;
-#endif
- if (exec_bfd == NULL)
- error (_("Can't find bfd architecture for corefile (need execfile)."));
-
- return bfd_get_arch (exec_bfd);
-}
-
-static char *
-default_gcore_target (void)
-{
- /* FIXME: This may only work for ELF targets. */
- if (exec_bfd == NULL)
- return NULL;
- else
- return bfd_get_target (exec_bfd);
-}
-
-/* Derive a reasonable stack segment by unwinding the target stack,
- and store its limits in *BOTTOM and *TOP. Return non-zero if
- successful. */
-
-static int
-derive_stack_segment (bfd_vma *bottom, bfd_vma *top)
-{
- struct frame_info *fi, *tmp_fi;
-
- gdb_assert (bottom);
- gdb_assert (top);
-
- /* Can't succeed without stack and registers. */
- if (!target_has_stack || !target_has_registers)
- return 0;
-
- /* Can't succeed without current frame. */
- fi = get_current_frame ();
- if (fi == NULL)
- return 0;
-
- /* Save frame pointer of TOS frame. */
- *top = get_frame_base (fi);
- /* If current stack pointer is more "inner", use that instead. */
- if (INNER_THAN (read_sp (), *top))
- *top = read_sp ();
-
- /* Find prev-most frame. */
- while ((tmp_fi = get_prev_frame (fi)) != NULL)
- fi = tmp_fi;
-
- /* Save frame pointer of prev-most frame. */
- *bottom = get_frame_base (fi);
-
- /* Now canonicalize their order, so that BOTTOM is a lower address
- (as opposed to a lower stack frame). */
- if (*bottom > *top)
- {
- bfd_vma tmp_vma;
-
- tmp_vma = *top;
- *top = *bottom;
- *bottom = tmp_vma;
- }
-
- return 1;
-}
-
-/* Derive a reasonable heap segment for ABFD by looking at sbrk and
- the static data sections. Store its limits in *BOTTOM and *TOP.
- Return non-zero if successful. */
-
-static int
-derive_heap_segment (bfd *abfd, bfd_vma *bottom, bfd_vma *top)
-{
- bfd_vma top_of_data_memory = 0;
- bfd_vma top_of_heap = 0;
- bfd_size_type sec_size;
- struct value *zero, *sbrk;
- bfd_vma sec_vaddr;
- asection *sec;
-
- gdb_assert (bottom);
- gdb_assert (top);
-
- /* This function depends on being able to call a function in the
- inferior. */
- if (!target_has_execution)
- return 0;
-
- /* The following code assumes that the link map is arranged as
- follows (low to high addresses):
-
- ---------------------------------
- | text sections |
- ---------------------------------
- | data sections (including bss) |
- ---------------------------------
- | heap |
- --------------------------------- */
-
- for (sec = abfd->sections; sec; sec = sec->next)
- {
- if (bfd_get_section_flags (abfd, sec) & SEC_DATA
- || strcmp (".bss", bfd_section_name (abfd, sec)) == 0)
- {
- sec_vaddr = bfd_get_section_vma (abfd, sec);
- sec_size = bfd_get_section_size (sec);
- if (sec_vaddr + sec_size > top_of_data_memory)
- top_of_data_memory = sec_vaddr + sec_size;
- }
- }
-
- /* Now get the top-of-heap by calling sbrk in the inferior. */
- if (lookup_minimal_symbol ("sbrk", NULL, NULL) != NULL)
- {
- sbrk = find_function_in_inferior ("sbrk");
- if (sbrk == NULL)
- return 0;
- }
- else if (lookup_minimal_symbol ("_sbrk", NULL, NULL) != NULL)
- {
- sbrk = find_function_in_inferior ("_sbrk");
- if (sbrk == NULL)
- return 0;
- }
- else
- return 0;
-
- zero = value_from_longest (builtin_type_int, 0);
- gdb_assert (zero);
- sbrk = call_function_by_hand (sbrk, 1, &zero);
- if (sbrk == NULL)
- return 0;
- top_of_heap = value_as_long (sbrk);
-
- /* Return results. */
- if (top_of_heap > top_of_data_memory)
- {
- *bottom = top_of_data_memory;
- *top = top_of_heap;
- return 1;
- }
-
- /* No additional heap space needs to be saved. */
- return 0;
-}
-
-static void
-make_output_phdrs (bfd *obfd, asection *osec, void *ignored)
-{
- int p_flags = 0;
- int p_type;
-
- /* FIXME: these constants may only be applicable for ELF. */
- if (strncmp (bfd_section_name (obfd, osec), "load", 4) == 0)
- p_type = PT_LOAD;
- else
- p_type = PT_NOTE;
-
- p_flags |= PF_R; /* Segment is readable. */
- if (!(bfd_get_section_flags (obfd, osec) & SEC_READONLY))
- p_flags |= PF_W; /* Segment is writable. */
- if (bfd_get_section_flags (obfd, osec) & SEC_CODE)
- p_flags |= PF_X; /* Segment is executable. */
-
- bfd_record_phdr (obfd, p_type, 1, p_flags, 0, 0, 0, 0, 1, &osec);
-}
-
-static int
-gcore_create_callback (CORE_ADDR vaddr, unsigned long size,
- int read, int write, int exec, void *data)
-{
- bfd *obfd = data;
- asection *osec;
- flagword flags = SEC_ALLOC | SEC_HAS_CONTENTS | SEC_LOAD;
-
- /* If the memory segment has no permissions set, ignore it, otherwise
- when we later try to access it for read/write, we'll get an error
- or jam the kernel. */
- if (read == 0 && write == 0 && exec == 0)
- {
- if (info_verbose)
- {
- fprintf_filtered (gdb_stdout, "Ignore segment, %s bytes at 0x%s\n",
- paddr_d (size), paddr_nz (vaddr));
- }
-
- return 0;
- }
-
- if (write == 0)
- {
- /* See if this region of memory lies inside a known file on disk.
- If so, we can avoid copying its contents by clearing SEC_LOAD. */
- struct objfile *objfile;
- struct obj_section *objsec;
-
- ALL_OBJSECTIONS (objfile, objsec)
- {
- bfd *abfd = objfile->obfd;
- asection *asec = objsec->the_bfd_section;
- bfd_vma align = (bfd_vma) 1 << bfd_get_section_alignment (abfd,
- asec);
- bfd_vma start = objsec->addr & -align;
- bfd_vma end = (objsec->endaddr + align - 1) & -align;
- /* Match if either the entire memory region lies inside the
- section (i.e. a mapping covering some pages of a large
- segment) or the entire section lies inside the memory region
- (i.e. a mapping covering multiple small sections).
-
- This BFD was synthesized from reading target memory,
- we don't want to omit that. */
- if (((vaddr >= start && vaddr + size <= end)
- || (start >= vaddr && end <= vaddr + size))
- && !(bfd_get_file_flags (abfd) & BFD_IN_MEMORY))
- {
- flags &= ~SEC_LOAD;
- goto keep; /* break out of two nested for loops */
- }
- }
-
- keep:
- flags |= SEC_READONLY;
- }
-
- if (exec)
- flags |= SEC_CODE;
- else
- flags |= SEC_DATA;
-
- osec = bfd_make_section_anyway (obfd, "load");
- if (osec == NULL)
- {
- warning (_("Couldn't make gcore segment: %s"),
- bfd_errmsg (bfd_get_error ()));
- return 1;
- }
-
- if (info_verbose)
- {
- fprintf_filtered (gdb_stdout, "Save segment, %s bytes at 0x%s\n",
- paddr_d (size), paddr_nz (vaddr));
- }
-
- bfd_set_section_size (obfd, osec, size);
- bfd_set_section_vma (obfd, osec, vaddr);
- bfd_section_lma (obfd, osec) = 0; /* ??? bfd_set_section_lma? */
- bfd_set_section_flags (obfd, osec, flags);
- return 0;
-}
-
-static int
-objfile_find_memory_regions (int (*func) (CORE_ADDR, unsigned long,
- int, int, int, void *),
- void *obfd)
-{
- /* Use objfile data to create memory sections. */
- struct objfile *objfile;
- struct obj_section *objsec;
- bfd_vma temp_bottom, temp_top;
-
- /* Call callback function for each objfile section. */
- ALL_OBJSECTIONS (objfile, objsec)
- {
- bfd *ibfd = objfile->obfd;
- asection *isec = objsec->the_bfd_section;
- flagword flags = bfd_get_section_flags (ibfd, isec);
- int ret;
-
- if ((flags & SEC_ALLOC) || (flags & SEC_LOAD))
- {
- int size = bfd_section_size (ibfd, isec);
- int ret;
-
- ret = (*func) (objsec->addr, bfd_section_size (ibfd, isec),
- 1, /* All sections will be readable. */
- (flags & SEC_READONLY) == 0, /* Writable. */
- (flags & SEC_CODE) != 0, /* Executable. */
- obfd);
- if (ret != 0)
- return ret;
- }
- }
-
- /* Make a stack segment. */
- if (derive_stack_segment (&temp_bottom, &temp_top))
- (*func) (temp_bottom, temp_top - temp_bottom,
- 1, /* Stack section will be readable. */
- 1, /* Stack section will be writable. */
- 0, /* Stack section will not be executable. */
- obfd);
-
- /* Make a heap segment. */
- if (derive_heap_segment (exec_bfd, &temp_bottom, &temp_top))
- (*func) (temp_bottom, temp_top - temp_bottom,
- 1, /* Heap section will be readable. */
- 1, /* Heap section will be writable. */
- 0, /* Heap section will not be executable. */
- obfd);
-
- return 0;
-}
-
-static void
-gcore_copy_callback (bfd *obfd, asection *osec, void *ignored)
-{
- bfd_size_type size = bfd_section_size (obfd, osec);
- struct cleanup *old_chain = NULL;
- void *memhunk;
-
- /* Read-only sections are marked; we don't have to copy their contents. */
- if ((bfd_get_section_flags (obfd, osec) & SEC_LOAD) == 0)
- return;
-
- /* Only interested in "load" sections. */
- if (strncmp ("load", bfd_section_name (obfd, osec), 4) != 0)
- return;
-
- memhunk = xmalloc (size);
- /* ??? This is crap since xmalloc should never return NULL. */
- if (memhunk == NULL)
- error (_("Not enough memory to create corefile."));
- old_chain = make_cleanup (xfree, memhunk);
-
- if (target_read_memory (bfd_section_vma (obfd, osec),
- memhunk, size) != 0)
- warning (_("Memory read failed for corefile section, %s bytes at 0x%s."),
- paddr_d (size), paddr (bfd_section_vma (obfd, osec)));
- if (!bfd_set_section_contents (obfd, osec, memhunk, 0, size))
- warning (_("Failed to write corefile contents (%s)."),
- bfd_errmsg (bfd_get_error ()));
-
- do_cleanups (old_chain); /* Frees MEMHUNK. */
-}
-
-static int
-gcore_memory_sections (bfd *obfd)
-{
- if (target_find_memory_regions (gcore_create_callback, obfd) != 0)
- return 0; /* FIXME: error return/msg? */
-
- /* Record phdrs for section-to-segment mapping. */
- bfd_map_over_sections (obfd, make_output_phdrs, NULL);
-
- /* Copy memory region contents. */
- bfd_map_over_sections (obfd, gcore_copy_callback, NULL);
-
- return 1;
+ target_write_core_file (ptid_get_pid (inferior_ptid), corefilename);
}
void
@@ -496,5 +58,4 @@ Save a core file with the current state
Argument is optional filename. Default filename is 'core.<process_id>'."));
add_com_alias ("gcore", "generate-core-file", class_files, 1);
- exec_set_find_memory_regions (objfile_find_memory_regions);
}
Index: Makefile.in
===================================================================
RCS file: /cvs/src/src/gdb/Makefile.in,v
retrieving revision 1.772
diff -u -p -r1.772 Makefile.in
--- Makefile.in 4 Jan 2006 19:34:58 -0000 1.772
+++ Makefile.in 10 Jan 2006 16:01:52 -0000
@@ -940,7 +940,7 @@ COMMON_OBS = $(DEPFILES) $(CONFIG_OBS) $
reggroups.o regset.o \
trad-frame.o \
tramp-frame.o \
- solib.o solib-null.o
+ solib.o solib-null.o gcore.o
TSOBS = inflow.o
@@ -1971,9 +1971,9 @@ f-typeprint.o: f-typeprint.c $(defs_h) $
f-valprint.o: f-valprint.c $(defs_h) $(gdb_string_h) $(symtab_h) \
$(gdbtypes_h) $(expression_h) $(value_h) $(valprint_h) $(language_h) \
$(f_lang_h) $(frame_h) $(gdbcore_h) $(command_h) $(block_h)
-gcore.o: gcore.c $(defs_h) $(elf_bfd_h) $(infcall_h) $(inferior_h) \
- $(gdbcore_h) $(objfiles_h) $(symfile_h) $(cli_decode_h) \
- $(gdb_assert_h)
+gcore.o: gcore.c $(defs_h) $(inferior_h) $(target_h) $(command_h)
+gcore-elf.o: gcore-elf.c $(defs_h) $(elf_bfd_h) $(infcall_h) \
+ $(inferior_h) $(gdbcore_h) $(gdb_assert_h)
gdbarch.o: gdbarch.c $(defs_h) $(arch_utils_h) $(gdbcmd_h) $(inferior_h) \
$(symcat_h) $(floatformat_h) $(gdb_assert_h) $(gdb_string_h) \
$(gdb_events_h) $(reggroups_h) $(osabi_h) $(gdb_obstack_h)
Index: inf-ttrace.c
===================================================================
RCS file: /cvs/src/src/gdb/inf-ttrace.c,v
retrieving revision 1.18
diff -u -p -r1.18 inf-ttrace.c
--- inf-ttrace.c 23 Dec 2005 20:51:35 -0000 1.18
+++ inf-ttrace.c 10 Jan 2006 16:02:46 -0000
@@ -1116,6 +1116,31 @@ inf_ttrace_pid_to_str (ptid_t ptid)
return normal_pid_to_str (ptid);
}
+
+static void
+inf_ttrace_write_core_file (int pid, const char *filename)
+{
+ int status;
+ char default_filename[64];
+
+ /* Certain versions of HP/UX do not support the specification of the core
+ file name. To work around this, we simply do not specify it, let
+ the kernel dump the core file in the default file name, and we then
+ rename it. */
+ status = ttrace (TT_PROC_CORE, pid, 0, 0, 0, 0);
+
+ if (status == -1)
+ perror_with_name ("ttrace");
+
+ sprintf (default_filename, "core.%d", pid);
+ status = rename (default_filename, filename);
+
+ if (status != 0)
+ perror_with_name ("Unable to save core file");
+ else
+ fprintf_filtered (gdb_stdout, "Saved corefile %s\n", filename);
+}
+
\f
struct target_ops *
@@ -1140,6 +1165,7 @@ inf_ttrace_target (void)
t->to_mourn_inferior = inf_ttrace_mourn_inferior;
t->to_thread_alive = inf_ttrace_thread_alive;
t->to_pid_to_str = inf_ttrace_pid_to_str;
+ t->to_write_core_file = inf_ttrace_write_core_file;
t->to_xfer_partial = inf_ttrace_xfer_partial;
ttrace_ops_hack = t;
Index: config/alpha/alpha-linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/alpha/alpha-linux.mh,v
retrieving revision 1.19
diff -u -p -r1.19 alpha-linux.mh
--- config/alpha/alpha-linux.mh 4 Jan 2006 19:33:09 -0000 1.19
+++ config/alpha/alpha-linux.mh 10 Jan 2006 16:03:38 -0000
@@ -1,7 +1,7 @@
# Host: Little-endian Alpha running Linux
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o corelow.o alpha-nat.o alpha-linux-nat.o \
- fork-child.o proc-service.o linux-thread-db.o gcore.o \
+ fork-child.o proc-service.o linux-thread-db.o gcore-elf.o \
linux-nat.o linux-fork.o
# The dynamically loaded libthread_db needs access to symbols in the
Index: config/alpha/fbsd.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/alpha/fbsd.mh,v
retrieving revision 1.7
diff -u -p -r1.7 fbsd.mh
--- config/alpha/fbsd.mh 5 Mar 2005 14:36:10 -0000 1.7
+++ config/alpha/fbsd.mh 10 Jan 2006 16:03:38 -0000
@@ -1,6 +1,6 @@
# Host: FreeBSD/alpha
NATDEPFILES= fork-child.o inf-ptrace.o \
fbsd-nat.o alphabsd-nat.o \
- gcore.o solib.o solib-svr4.o solib-legacy.o \
+ gcore-elf.o solib.o solib-svr4.o solib-legacy.o \
corelow.o core-regset.o
NAT_FILE= nm-fbsd.h
Index: config/arm/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/arm/linux.mh,v
retrieving revision 1.17
diff -u -p -r1.17 linux.mh
--- config/arm/linux.mh 4 Jan 2006 19:33:09 -0000 1.17
+++ config/arm/linux.mh 10 Jan 2006 16:03:38 -0000
@@ -2,7 +2,7 @@
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o corelow.o \
- core-regset.o arm-linux-nat.o gcore.o \
+ core-regset.o arm-linux-nat.o gcore-elf.o \
proc-service.o linux-thread-db.o linux-nat.o linux-fork.o
LOADLIBES= -ldl -rdynamic
Index: config/i386/fbsd.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/fbsd.mh,v
retrieving revision 1.19
diff -u -p -r1.19 fbsd.mh
--- config/i386/fbsd.mh 1 Oct 2004 17:26:12 -0000 1.19
+++ config/i386/fbsd.mh 10 Jan 2006 16:03:38 -0000
@@ -1,7 +1,7 @@
# Host: FreeBSD/i386
NATDEPFILES= fork-child.o inf-ptrace.o \
fbsd-nat.o i386-nat.o i386bsd-nat.o i386fbsd-nat.o \
- gcore.o bsd-kvm.o
+ gcore-elf.o bsd-kvm.o
NAT_FILE= nm-fbsd.h
LOADLIBES= -lkvm
Index: config/i386/fbsd64.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/fbsd64.mh,v
retrieving revision 1.11
diff -u -p -r1.11 fbsd64.mh
--- config/i386/fbsd64.mh 1 Oct 2004 17:26:13 -0000 1.11
+++ config/i386/fbsd64.mh 10 Jan 2006 16:03:38 -0000
@@ -1,6 +1,6 @@
# Host: FreeBSD/amd64
NATDEPFILES= fork-child.o inf-ptrace.o \
fbsd-nat.o amd64-nat.o amd64bsd-nat.o amd64fbsd-nat.o \
- gcore.o bsd-kvm.o
+ gcore-elf.o bsd-kvm.o
LOADLIBES= -lkvm
Index: config/i386/i386sol2.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/i386sol2.mh,v
retrieving revision 1.11
diff -u -p -r1.11 i386sol2.mh
--- config/i386/i386sol2.mh 31 Oct 2004 15:35:08 -0000 1.11
+++ config/i386/i386sol2.mh 10 Jan 2006 16:03:38 -0000
@@ -1,4 +1,4 @@
# Host: Solaris x86
NATDEPFILES= fork-child.o i386v4-nat.o i386-sol2-nat.o \
- procfs.o proc-api.o proc-events.o proc-flags.o proc-why.o gcore.o
+ procfs.o proc-api.o proc-events.o proc-flags.o proc-why.o gcore-elf.o
NAT_FILE= nm-i386sol2.h
Index: config/i386/i386v42mp.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/i386v42mp.mh,v
retrieving revision 1.9
diff -u -p -r1.9 i386v42mp.mh
--- config/i386/i386v42mp.mh 5 Aug 2004 15:23:53 -0000 1.9
+++ config/i386/i386v42mp.mh 10 Jan 2006 16:03:38 -0000
@@ -11,6 +11,6 @@ NAT_FILE= nm-i386v42mp.h
# consequence, make considers subsequent tab-indented lines to be
# some sort of error.
NATDEPFILES= corelow.o core-regset.o fork-child.o i386v4-nat.o \
- gcore.o solib.o solib-svr4.o solib-legacy.o procfs.o proc-api.o \
+ gcore-elf.o solib.o solib-svr4.o solib-legacy.o procfs.o proc-api.o \
proc-events.o proc-flags.o proc-why.o uw-thread.o
Index: config/i386/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/linux.mh,v
retrieving revision 1.19
diff -u -p -r1.19 linux.mh
--- config/i386/linux.mh 4 Jan 2006 19:33:11 -0000 1.19
+++ config/i386/linux.mh 10 Jan 2006 16:03:38 -0000
@@ -3,7 +3,7 @@
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o corelow.o \
core-aout.o i386-nat.o i386-linux-nat.o \
- proc-service.o linux-thread-db.o gcore.o \
+ proc-service.o linux-thread-db.o gcore-elf.o \
linux-nat.o linux-fork.o
# The dynamically loaded libthread_db needs access to symbols in the
Index: config/i386/linux64.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/linux64.mh,v
retrieving revision 1.7
diff -u -p -r1.7 linux64.mh
--- config/i386/linux64.mh 4 Jan 2006 19:33:11 -0000 1.7
+++ config/i386/linux64.mh 10 Jan 2006 16:03:38 -0000
@@ -1,7 +1,7 @@
# Host: GNU/Linux x86-64
NATDEPFILES= inf-ptrace.o fork-child.o \
i386-nat.o amd64-nat.o amd64-linux-nat.o linux-nat.o \
- proc-service.o linux-thread-db.o gcore.o linux-fork.o
+ proc-service.o linux-thread-db.o gcore-elf.o linux-fork.o
NAT_FILE= nm-linux64.h
# The dynamically loaded libthread_db needs access to symbols in the
Index: config/i386/sol2-64.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/i386/sol2-64.mh,v
retrieving revision 1.1
diff -u -p -r1.1 sol2-64.mh
--- config/i386/sol2-64.mh 1 Nov 2004 21:05:56 -0000 1.1
+++ config/i386/sol2-64.mh 10 Jan 2006 16:03:39 -0000
@@ -1,4 +1,4 @@
# Host: Solaris x86_64
NATDEPFILES= fork-child.o amd64-nat.o i386v4-nat.o i386-sol2-nat.o \
- procfs.o proc-api.o proc-events.o proc-flags.o proc-why.o gcore.o
+ procfs.o proc-api.o proc-events.o proc-flags.o proc-why.o gcore-elf.o
NAT_FILE= nm-i386sol2.h
Index: config/ia64/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/ia64/linux.mh,v
retrieving revision 1.20
diff -u -p -r1.20 linux.mh
--- config/ia64/linux.mh 4 Jan 2006 19:33:11 -0000 1.20
+++ config/ia64/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -1,7 +1,7 @@
# Host: Intel IA-64 running GNU/Linux
NAT_FILE= nm-linux.h
-NATDEPFILES= inf-ptrace.o fork-child.o corelow.o gcore.o \
+NATDEPFILES= inf-ptrace.o fork-child.o corelow.o gcore-elf.o \
core-aout.o core-regset.o ia64-linux-nat.o \
proc-service.o linux-thread-db.o linux-nat.o linux-fork.o
Index: config/m32r/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/m32r/linux.mh,v
retrieving revision 1.4
diff -u -p -r1.4 linux.mh
--- config/m32r/linux.mh 4 Jan 2006 19:33:11 -0000 1.4
+++ config/m32r/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -2,7 +2,7 @@
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o corelow.o \
- m32r-linux-nat.o gcore.o proc-service.o linux-thread-db.o \
+ m32r-linux-nat.o gcore-elf.o proc-service.o linux-thread-db.o \
linux-nat.o linux-fork.o
LOADLIBES= -ldl -rdynamic
Index: config/m68k/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/m68k/linux.mh,v
retrieving revision 1.18
diff -u -p -r1.18 linux.mh
--- config/m68k/linux.mh 4 Jan 2006 19:33:11 -0000 1.18
+++ config/m68k/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -2,7 +2,7 @@
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o \
- corelow.o core-aout.o m68klinux-nat.o gcore.o \
+ corelow.o core-aout.o m68klinux-nat.o gcore-elf.o \
proc-service.o linux-thread-db.o linux-nat.o linux-fork.o
# The dynamically loaded libthread_db needs access to symbols in the
Index: config/mips/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/mips/linux.mh,v
retrieving revision 1.11
diff -u -p -r1.11 linux.mh
--- config/mips/linux.mh 4 Jan 2006 19:33:12 -0000 1.11
+++ config/mips/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -1,7 +1,7 @@
# Host: Linux/MIPS
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o mips-linux-nat.o \
- linux-thread-db.o proc-service.o gcore.o \
+ linux-thread-db.o proc-service.o gcore-elf.o \
linux-nat.o linux-fork.o
LOADLIBES = -ldl -rdynamic
Index: config/pa/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/pa/linux.mh,v
retrieving revision 1.7
diff -u -p -r1.7 linux.mh
--- config/pa/linux.mh 4 Jan 2006 19:33:12 -0000 1.7
+++ config/pa/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -1,7 +1,7 @@
# Host: Hewlett-Packard PA-RISC machine, running Linux
XDEPFILES=
NAT_FILE= nm-linux.h
-NATDEPFILES= inf-ptrace.o fork-child.o corelow.o gcore.o \
+NATDEPFILES= inf-ptrace.o fork-child.o corelow.o gcore-elf.o \
core-regset.o hppa-linux-nat.o \
proc-service.o linux-thread-db.o linux-nat.o linux-fork.o
Index: config/powerpc/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/powerpc/linux.mh,v
retrieving revision 1.20
diff -u -p -r1.20 linux.mh
--- config/powerpc/linux.mh 4 Jan 2006 19:33:12 -0000 1.20
+++ config/powerpc/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -5,6 +5,6 @@ XM_CLIBS=
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o \
ppc-linux-nat.o proc-service.o linux-thread-db.o \
- gcore.o linux-nat.o linux-fork.o
+ gcore-elf.o linux-nat.o linux-fork.o
LOADLIBES = -ldl -rdynamic
Index: config/powerpc/ppc64-linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/powerpc/ppc64-linux.mh,v
retrieving revision 1.9
diff -u -p -r1.9 ppc64-linux.mh
--- config/powerpc/ppc64-linux.mh 4 Jan 2006 19:33:12 -0000 1.9
+++ config/powerpc/ppc64-linux.mh 10 Jan 2006 16:03:39 -0000
@@ -5,7 +5,7 @@ XM_CLIBS=
NAT_FILE= nm-ppc64-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o \
ppc-linux-nat.o proc-service.o linux-thread-db.o \
- gcore.o linux-nat.o linux-fork.o
+ gcore-elf.o linux-nat.o linux-fork.o
# The PowerPC has severe limitations on TOC size, and uses them even
# for non-PIC code. GDB overflows those tables when compiling with
Index: config/s390/s390.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/s390/s390.mh,v
retrieving revision 1.13
diff -u -p -r1.13 s390.mh
--- config/s390/s390.mh 10 Sep 2005 18:11:13 -0000 1.13
+++ config/s390/s390.mh 10 Jan 2006 16:03:39 -0000
@@ -1,5 +1,5 @@
# Host: S390, running Linux
NAT_FILE= nm-linux.h
NATDEPFILES= inf-ptrace.o fork-child.o corelow.o s390-nat.o \
- gcore.o linux-thread-db.o proc-service.o linux-nat.o
+ gcore-elf.o linux-thread-db.o proc-service.o linux-nat.o
LOADLIBES = -ldl -rdynamic
Index: config/sparc/fbsd.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/sparc/fbsd.mh,v
retrieving revision 1.9
diff -u -p -r1.9 fbsd.mh
--- config/sparc/fbsd.mh 15 Jul 2005 20:10:19 -0000 1.9
+++ config/sparc/fbsd.mh 10 Jan 2006 16:03:39 -0000
@@ -1,6 +1,6 @@
# Host: FreeBSD/sparc64
NATDEPFILES= fork-child.o inf-ptrace.o \
fbsd-nat.o sparc-nat.o sparc64-nat.o sparc64fbsd-nat.o \
- gcore.o bsd-kvm.o
+ gcore-elf.o bsd-kvm.o
LOADLIBES= -lkvm
Index: config/sparc/linux.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/sparc/linux.mh,v
retrieving revision 1.17
diff -u -p -r1.17 linux.mh
--- config/sparc/linux.mh 4 Jan 2006 19:33:12 -0000 1.17
+++ config/sparc/linux.mh 10 Jan 2006 16:03:39 -0000
@@ -3,7 +3,7 @@ NAT_FILE= nm-linux.h
NATDEPFILES= sparc-nat.o sparc-sol2-nat.o sparc-linux-nat.o \
corelow.o core-regset.o fork-child.o inf-ptrace.o \
proc-service.o linux-thread-db.o \
- gcore.o linux-nat.o linux-fork.o
+ gcore-elf.o linux-nat.o linux-fork.o
# The dynamically loaded libthread_db needs access to symbols in the
# gdb executable.
Index: config/sparc/linux64.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/sparc/linux64.mh,v
retrieving revision 1.8
diff -u -p -r1.8 linux64.mh
--- config/sparc/linux64.mh 4 Jan 2006 19:33:13 -0000 1.8
+++ config/sparc/linux64.mh 10 Jan 2006 16:03:39 -0000
@@ -4,7 +4,7 @@ NATDEPFILES= sparc-nat.o sparc64-nat.o s
corelow.o core-regset.o \
fork-child.o inf-ptrace.o \
proc-service.o linux-thread-db.o \
- gcore.o linux-nat.o linux-fork.o
+ gcore-elf.o linux-nat.o linux-fork.o
# The dynamically loaded libthread_db needs access to symbols in the
# gdb executable.
Index: config/sparc/sol2.mh
===================================================================
RCS file: /cvs/src/src/gdb/config/sparc/sol2.mh,v
retrieving revision 1.3
diff -u -p -r1.3 sol2.mh
--- config/sparc/sol2.mh 7 Apr 2004 00:15:12 -0000 1.3
+++ config/sparc/sol2.mh 10 Jan 2006 16:03:39 -0000
@@ -1,5 +1,5 @@
# Host: Solaris SPARC & UltraSPARC
NAT_FILE= nm-sol2.h
NATDEPFILES= sparc-sol2-nat.o \
- corelow.o core-regset.o fork-child.o gcore.o \
+ corelow.o core-regset.o fork-child.o gcore-elf.o \
procfs.o proc-api.o proc-events.o proc-flags.o proc-why.o
[-- Attachment #3: gcore-elf.c --]
[-- Type: text/plain, Size: 13846 bytes --]
/* Generate a core file for the inferior process.
Copyright (C) 2006 Free Software Foundation, Inc.
This file is part of GDB.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor,
Boston, MA 02110-1301, USA. */
#include "defs.h"
#include "elf-bfd.h"
#include "infcall.h"
#include "inferior.h"
#include "gdbcore.h"
#include "objfiles.h"
#include "gdb_assert.h"
static char *default_gcore_target (void);
static enum bfd_architecture default_gcore_arch (void);
static unsigned long default_gcore_mach (void);
static int gcore_memory_sections (bfd *);
static void gcore_elf_write_core_file (int pid, const char *corefilename);
/* Generate a core file from the inferior process. */
static void
gcore_elf_write_core_file (int pid, const char *corefilename)
{
struct cleanup *old_chain;
asection *note_sec = NULL;
bfd *obfd;
void *note_data = NULL;
int note_size = 0;
if (info_verbose)
fprintf_filtered (gdb_stdout,
"Opening corefile '%s' for output.\n", corefilename);
/* Open the output file. */
obfd = bfd_openw (corefilename, default_gcore_target ());
if (!obfd)
error (_("Failed to open '%s' for output."), corefilename);
/* Need a cleanup that will close the file (FIXME: delete it?). */
old_chain = make_cleanup_bfd_close (obfd);
bfd_set_format (obfd, bfd_core);
bfd_set_arch_mach (obfd, default_gcore_arch (), default_gcore_mach ());
/* An external target method must build the notes section. */
note_data = target_make_corefile_notes (obfd, ¬e_size);
/* Create the note section. */
if (note_data != NULL && note_size != 0)
{
note_sec = bfd_make_section_anyway (obfd, "note0");
if (note_sec == NULL)
error (_("Failed to create 'note' section for corefile: %s"),
bfd_errmsg (bfd_get_error ()));
bfd_set_section_vma (obfd, note_sec, 0);
bfd_set_section_flags (obfd, note_sec,
SEC_HAS_CONTENTS | SEC_READONLY | SEC_ALLOC);
bfd_set_section_alignment (obfd, note_sec, 0);
bfd_set_section_size (obfd, note_sec, note_size);
}
/* Now create the memory/load sections. */
if (gcore_memory_sections (obfd) == 0)
error (_("gcore: failed to get corefile memory sections from target."));
/* Write out the contents of the note section. */
if (note_data != NULL && note_size != 0)
{
if (!bfd_set_section_contents (obfd, note_sec, note_data, 0, note_size))
warning (_("writing note section (%s)"), bfd_errmsg (bfd_get_error ()));
}
/* Succeeded. */
fprintf_filtered (gdb_stdout, "Saved corefile %s\n", corefilename);
/* Clean-ups will close the output file and free malloc memory. */
do_cleanups (old_chain);
return;
}
static unsigned long
default_gcore_mach (void)
{
#if 1 /* See if this even matters... */
return 0;
#else
#ifdef TARGET_ARCHITECTURE
const struct bfd_arch_info *bfdarch = TARGET_ARCHITECTURE;
if (bfdarch != NULL)
return bfdarch->mach;
#endif /* TARGET_ARCHITECTURE */
if (exec_bfd == NULL)
error (_("Can't find default bfd machine type (need execfile)."));
return bfd_get_mach (exec_bfd);
#endif /* 1 */
}
static enum bfd_architecture
default_gcore_arch (void)
{
#ifdef TARGET_ARCHITECTURE
const struct bfd_arch_info * bfdarch = TARGET_ARCHITECTURE;
if (bfdarch != NULL)
return bfdarch->arch;
#endif
if (exec_bfd == NULL)
error (_("Can't find bfd architecture for corefile (need execfile)."));
return bfd_get_arch (exec_bfd);
}
static char *
default_gcore_target (void)
{
/* FIXME: This may only work for ELF targets. */
if (exec_bfd == NULL)
return NULL;
else
return bfd_get_target (exec_bfd);
}
/* Derive a reasonable stack segment by unwinding the target stack,
and store its limits in *BOTTOM and *TOP. Return non-zero if
successful. */
static int
derive_stack_segment (bfd_vma *bottom, bfd_vma *top)
{
struct frame_info *fi, *tmp_fi;
gdb_assert (bottom);
gdb_assert (top);
/* Can't succeed without stack and registers. */
if (!target_has_stack || !target_has_registers)
return 0;
/* Can't succeed without current frame. */
fi = get_current_frame ();
if (fi == NULL)
return 0;
/* Save frame pointer of TOS frame. */
*top = get_frame_base (fi);
/* If current stack pointer is more "inner", use that instead. */
if (INNER_THAN (read_sp (), *top))
*top = read_sp ();
/* Find prev-most frame. */
while ((tmp_fi = get_prev_frame (fi)) != NULL)
fi = tmp_fi;
/* Save frame pointer of prev-most frame. */
*bottom = get_frame_base (fi);
/* Now canonicalize their order, so that BOTTOM is a lower address
(as opposed to a lower stack frame). */
if (*bottom > *top)
{
bfd_vma tmp_vma;
tmp_vma = *top;
*top = *bottom;
*bottom = tmp_vma;
}
return 1;
}
/* Derive a reasonable heap segment for ABFD by looking at sbrk and
the static data sections. Store its limits in *BOTTOM and *TOP.
Return non-zero if successful. */
static int
derive_heap_segment (bfd *abfd, bfd_vma *bottom, bfd_vma *top)
{
bfd_vma top_of_data_memory = 0;
bfd_vma top_of_heap = 0;
bfd_size_type sec_size;
struct value *zero, *sbrk;
bfd_vma sec_vaddr;
asection *sec;
gdb_assert (bottom);
gdb_assert (top);
/* This function depends on being able to call a function in the
inferior. */
if (!target_has_execution)
return 0;
/* The following code assumes that the link map is arranged as
follows (low to high addresses):
---------------------------------
| text sections |
---------------------------------
| data sections (including bss) |
---------------------------------
| heap |
--------------------------------- */
for (sec = abfd->sections; sec; sec = sec->next)
{
if (bfd_get_section_flags (abfd, sec) & SEC_DATA
|| strcmp (".bss", bfd_section_name (abfd, sec)) == 0)
{
sec_vaddr = bfd_get_section_vma (abfd, sec);
sec_size = bfd_get_section_size (sec);
if (sec_vaddr + sec_size > top_of_data_memory)
top_of_data_memory = sec_vaddr + sec_size;
}
}
/* Now get the top-of-heap by calling sbrk in the inferior. */
if (lookup_minimal_symbol ("sbrk", NULL, NULL) != NULL)
{
sbrk = find_function_in_inferior ("sbrk");
if (sbrk == NULL)
return 0;
}
else if (lookup_minimal_symbol ("_sbrk", NULL, NULL) != NULL)
{
sbrk = find_function_in_inferior ("_sbrk");
if (sbrk == NULL)
return 0;
}
else
return 0;
zero = value_from_longest (builtin_type_int, 0);
gdb_assert (zero);
sbrk = call_function_by_hand (sbrk, 1, &zero);
if (sbrk == NULL)
return 0;
top_of_heap = value_as_long (sbrk);
/* Return results. */
if (top_of_heap > top_of_data_memory)
{
*bottom = top_of_data_memory;
*top = top_of_heap;
return 1;
}
/* No additional heap space needs to be saved. */
return 0;
}
static void
make_output_phdrs (bfd *obfd, asection *osec, void *ignored)
{
int p_flags = 0;
int p_type;
/* FIXME: these constants may only be applicable for ELF. */
if (strncmp (bfd_section_name (obfd, osec), "load", 4) == 0)
p_type = PT_LOAD;
else
p_type = PT_NOTE;
p_flags |= PF_R; /* Segment is readable. */
if (!(bfd_get_section_flags (obfd, osec) & SEC_READONLY))
p_flags |= PF_W; /* Segment is writable. */
if (bfd_get_section_flags (obfd, osec) & SEC_CODE)
p_flags |= PF_X; /* Segment is executable. */
bfd_record_phdr (obfd, p_type, 1, p_flags, 0, 0, 0, 0, 1, &osec);
}
static int
gcore_create_callback (CORE_ADDR vaddr, unsigned long size,
int read, int write, int exec, void *data)
{
bfd *obfd = data;
asection *osec;
flagword flags = SEC_ALLOC | SEC_HAS_CONTENTS | SEC_LOAD;
/* If the memory segment has no permissions set, ignore it, otherwise
when we later try to access it for read/write, we'll get an error
or jam the kernel. */
if (read == 0 && write == 0 && exec == 0)
{
if (info_verbose)
{
fprintf_filtered (gdb_stdout, "Ignore segment, %s bytes at 0x%s\n",
paddr_d (size), paddr_nz (vaddr));
}
return 0;
}
if (write == 0)
{
/* See if this region of memory lies inside a known file on disk.
If so, we can avoid copying its contents by clearing SEC_LOAD. */
struct objfile *objfile;
struct obj_section *objsec;
ALL_OBJSECTIONS (objfile, objsec)
{
bfd *abfd = objfile->obfd;
asection *asec = objsec->the_bfd_section;
bfd_vma align = (bfd_vma) 1 << bfd_get_section_alignment (abfd,
asec);
bfd_vma start = objsec->addr & -align;
bfd_vma end = (objsec->endaddr + align - 1) & -align;
/* Match if either the entire memory region lies inside the
section (i.e. a mapping covering some pages of a large
segment) or the entire section lies inside the memory region
(i.e. a mapping covering multiple small sections).
This BFD was synthesized from reading target memory,
we don't want to omit that. */
if (((vaddr >= start && vaddr + size <= end)
|| (start >= vaddr && end <= vaddr + size))
&& !(bfd_get_file_flags (abfd) & BFD_IN_MEMORY))
{
flags &= ~SEC_LOAD;
goto keep; /* break out of two nested for loops */
}
}
keep:
flags |= SEC_READONLY;
}
if (exec)
flags |= SEC_CODE;
else
flags |= SEC_DATA;
osec = bfd_make_section_anyway (obfd, "load");
if (osec == NULL)
{
warning (_("Couldn't make gcore segment: %s"),
bfd_errmsg (bfd_get_error ()));
return 1;
}
if (info_verbose)
{
fprintf_filtered (gdb_stdout, "Save segment, %s bytes at 0x%s\n",
paddr_d (size), paddr_nz (vaddr));
}
bfd_set_section_size (obfd, osec, size);
bfd_set_section_vma (obfd, osec, vaddr);
bfd_section_lma (obfd, osec) = 0; /* ??? bfd_set_section_lma? */
bfd_set_section_flags (obfd, osec, flags);
return 0;
}
static int
objfile_find_memory_regions (int (*func) (CORE_ADDR, unsigned long,
int, int, int, void *),
void *obfd)
{
/* Use objfile data to create memory sections. */
struct objfile *objfile;
struct obj_section *objsec;
bfd_vma temp_bottom, temp_top;
/* Call callback function for each objfile section. */
ALL_OBJSECTIONS (objfile, objsec)
{
bfd *ibfd = objfile->obfd;
asection *isec = objsec->the_bfd_section;
flagword flags = bfd_get_section_flags (ibfd, isec);
int ret;
if ((flags & SEC_ALLOC) || (flags & SEC_LOAD))
{
int size = bfd_section_size (ibfd, isec);
int ret;
ret = (*func) (objsec->addr, bfd_section_size (ibfd, isec),
1, /* All sections will be readable. */
(flags & SEC_READONLY) == 0, /* Writable. */
(flags & SEC_CODE) != 0, /* Executable. */
obfd);
if (ret != 0)
return ret;
}
}
/* Make a stack segment. */
if (derive_stack_segment (&temp_bottom, &temp_top))
(*func) (temp_bottom, temp_top - temp_bottom,
1, /* Stack section will be readable. */
1, /* Stack section will be writable. */
0, /* Stack section will not be executable. */
obfd);
/* Make a heap segment. */
if (derive_heap_segment (exec_bfd, &temp_bottom, &temp_top))
(*func) (temp_bottom, temp_top - temp_bottom,
1, /* Heap section will be readable. */
1, /* Heap section will be writable. */
0, /* Heap section will not be executable. */
obfd);
return 0;
}
static void
gcore_copy_callback (bfd *obfd, asection *osec, void *ignored)
{
bfd_size_type size = bfd_section_size (obfd, osec);
struct cleanup *old_chain = NULL;
void *memhunk;
/* Read-only sections are marked; we don't have to copy their contents. */
if ((bfd_get_section_flags (obfd, osec) & SEC_LOAD) == 0)
return;
/* Only interested in "load" sections. */
if (strncmp ("load", bfd_section_name (obfd, osec), 4) != 0)
return;
memhunk = xmalloc (size);
/* ??? This is crap since xmalloc should never return NULL. */
if (memhunk == NULL)
error (_("Not enough memory to create corefile."));
old_chain = make_cleanup (xfree, memhunk);
if (target_read_memory (bfd_section_vma (obfd, osec),
memhunk, size) != 0)
warning (_("Memory read failed for corefile section, %s bytes at 0x%s."),
paddr_d (size), paddr (bfd_section_vma (obfd, osec)));
if (!bfd_set_section_contents (obfd, osec, memhunk, 0, size))
warning (_("Failed to write corefile contents (%s)."),
bfd_errmsg (bfd_get_error ()));
do_cleanups (old_chain); /* Frees MEMHUNK. */
}
static int
gcore_memory_sections (bfd *obfd)
{
if (target_find_memory_regions (gcore_create_callback, obfd) != 0)
return 0; /* FIXME: error return/msg? */
/* Record phdrs for section-to-segment mapping. */
bfd_map_over_sections (obfd, make_output_phdrs, NULL);
/* Copy memory region contents. */
bfd_map_over_sections (obfd, gcore_copy_callback, NULL);
return 1;
}
void
_initialize_gcore_elf (void)
{
exec_set_write_core_file (gcore_elf_write_core_file);
exec_set_find_memory_regions (objfile_find_memory_regions);
}
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RFA] implement gcore on hp/ux
2006-01-10 17:18 [RFA] implement gcore on hp/ux Joel Brobecker
@ 2006-01-10 20:29 ` Mark Kettenis
2006-01-11 3:59 ` Joel Brobecker
2006-01-12 5:55 ` Joel Brobecker
0 siblings, 2 replies; 6+ messages in thread
From: Mark Kettenis @ 2006-01-10 20:29 UTC (permalink / raw)
To: brobecker; +Cc: gdb-patches
> Date: Tue, 10 Jan 2006 21:17:52 +0400
> From: Joel Brobecker <brobecker@adacore.com>
> 2005-01-10 Joel Brobecker <brobecker@adacore.com>
>
> * target.h (target_ops): New method to_write_core_file.
> (target_write_core_file_p): New macro.
> (target_write_core_file): New macro.
> * target.c (update_current_target): Add heritance of to_write_core_file.
> * exec.c (exec_set_write_core_file): New function.
> * defs.h (exec_set_write_core_file): Add declaration.
> * gcore.c: Extract out the code writing the core file and move it
> to gcore-elf.c.
> * gcore-elf.c: New file, mostly extracted from gcore.o.
> * Makefile.in (COMMON_OBS): Add gcore.o.
> (gcore.o): Update dependencies.
> (gcore-elf.o): Add rule.
> * inf-ttrace (inf_ttrace_write_core_file): New function.
> (inf_ttrace_target): set the to_write_core_file method.
> * config/alpha/fbsd.mh (NATDEPFILES): Likewise.
> * config/arm/linux.mh (NATDEPFILES): Likewise.
> * config/i386/fbsd.mh (NATDEPFILES): Likewise.
> * config/i386/fbsd64.mh (NATDEPFILES): Likewise.
> * config/i386/i386sol2.mh (NATDEPFILES): Likewise.
> * config/i386/i386v42mp.mh (NATDEPFILES): Likewise.
> * config/i386/linux.mh (NATDEPFILES): Likewise.
> * config/i386/linux64.mh (NATDEPFILES): Likewise.
> * config/i386/sol2-64.mh (NATDEPFILES): Likewise.
> * config/ia64/linux.mh (NATDEPFILES): Likewise.
> * config/m32r/linux.mh (NATDEPFILES): Likewise.
> * config/m68k/linux.mh (NATDEPFILES): Likewise.
> * config/mips/linux.mh (NATDEPFILES): Likewise.
> * config/pa/linux.mh (NATDEPFILES): Likewise.
> * config/powerpc/linux.mh (NATDEPFILES): Likewise.
> * config/powerpc/ppc64-linux.mh (NATDEPFILES): Likewise.
> * config/s390/s390.mh (NATDEPFILES): Likewise.
> * config/sparc/fbsd.mh (NATDEPFILES): Likewise.
> * config/sparc/linux.mh (NATDEPFILES): Likewise.
> * config/sparc/linux64.mh (NATDEPFILES): Likewise.
> * config/sparc/sol2.mh (NATDEPFILES): Likewise.
>
> Tested on x86-linux and pa32-hpux. No regression on the linux
> side. On the hp/ux side, we get improvments from auxv.exp and
> gcore.exp now passes.
>
> Comments? Ok to apply?
I'm not sure I like the exec_set_write_core_file() hack. Targets
should really do this the proper way, initializing their target vector
properly. Can't you get rid of it by setting to_write_core_file in
procfs.c:init_ptoc_ops() and linux-nat.c:linux_target()?
The HP-UX code looks good, but could you please use xsnprintf()
instead of sprintf()?
Mark
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RFA] implement gcore on hp/ux
2006-01-10 20:29 ` Mark Kettenis
@ 2006-01-11 3:59 ` Joel Brobecker
2006-01-12 5:55 ` Joel Brobecker
1 sibling, 0 replies; 6+ messages in thread
From: Joel Brobecker @ 2006-01-11 3:59 UTC (permalink / raw)
To: Mark Kettenis; +Cc: gdb-patches
> I'm not sure I like the exec_set_write_core_file() hack. Targets
> should really do this the proper way, initializing their target vector
> properly. Can't you get rid of it by setting to_write_core_file in
> procfs.c:init_ptoc_ops() and linux-nat.c:linux_target()?
Yes, absolutely. I think I also need to do that for FreeBSD and
probably Solaris too (IIRC), right?
> The HP-UX code looks good, but could you please use xsnprintf()
> instead of sprintf()?
Oops, good catch. Will fix that as well.
Thanks for the review.
--
Joel
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RFA] implement gcore on hp/ux
2006-01-10 20:29 ` Mark Kettenis
2006-01-11 3:59 ` Joel Brobecker
@ 2006-01-12 5:55 ` Joel Brobecker
2006-01-12 20:44 ` Mark Kettenis
1 sibling, 1 reply; 6+ messages in thread
From: Joel Brobecker @ 2006-01-12 5:55 UTC (permalink / raw)
To: Mark Kettenis; +Cc: gdb-patches
Mark,
> I'm not sure I like the exec_set_write_core_file() hack. Targets
> should really do this the proper way, initializing their target vector
> properly. Can't you get rid of it by setting to_write_core_file in
> procfs.c:init_ptoc_ops() and linux-nat.c:linux_target()?
I ran into something I didn't expect. Doing the above for all linux
targets was easy; so was it for all freebsd targets. But I still have
to work on the following ones (names of mh files):
. i386/i386sol2
. i386/sol2-64
. sparc/sol2
. i386/i386v42mp
. s390/s390
All the above configurations currently include gcore.
I started working on i386sol2, and found out that the target vector
for this configuration seems to be the procfs one. In the rest of
this discussion, I will refer to "the" target vector as the one at
the process stratum.
For Linux and FreeBSD, it was easy because the nat files were always
creating their own target vector based on the ptrace one. The ptrace
target vector was never pushed on the target stack.
However, so far, the procfs target vector IS pushed on the target
stack, and no modification has been needed so far. But if we are
to continue in this direction, we will need to be able to modify
this target vector.
I was less and less convinced by the entire idea of using the target
vector, at least for such a minor extension to GDB just for HP/UX.
Perhaps a few lines of code duplication in inf-ttrace wouldn't be
so bad.
Except that it might be a good general cleanup for GDB. For instance,
I really like the model used by Linux for instance, where they use
a default target vector provided by inf-ptrace, and tailor it to
their needs. Would it make sense to modify the procfs support such
that it is no longer pushed itself on the stack, but instead used
as a tailorable template?
I'm sure we can find ways to scrub directly inside the procfs
target vector, but this would seem rather ugly to me.
The whole patch is becoming too big in my opinion to be submitted
in one go. If we agree on continuing on this path, then I will submit
one or more preparatory patches (such as making the procfs target vector
a template for instance), as needed.
Thanks,
--
Joel
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RFA] implement gcore on hp/ux
2006-01-12 5:55 ` Joel Brobecker
@ 2006-01-12 20:44 ` Mark Kettenis
2006-01-13 4:25 ` Michael Snyder
0 siblings, 1 reply; 6+ messages in thread
From: Mark Kettenis @ 2006-01-12 20:44 UTC (permalink / raw)
To: brobecker; +Cc: gdb-patches
> Date: Thu, 12 Jan 2006 09:55:29 +0400
> From: Joel Brobecker <brobecker@adacore.com>
>
> Mark,
>
> > I'm not sure I like the exec_set_write_core_file() hack. Targets
> > should really do this the proper way, initializing their target vector
> > properly. Can't you get rid of it by setting to_write_core_file in
> > procfs.c:init_ptoc_ops() and linux-nat.c:linux_target()?
>
> I ran into something I didn't expect. Doing the above for all linux
> targets was easy; so was it for all freebsd targets. But I still have
> to work on the following ones (names of mh files):
> . i386/i386sol2
> . i386/sol2-64
> . sparc/sol2
> . i386/i386v42mp
> . s390/s390
> All the above configurations currently include gcore.
>
> I started working on i386sol2, and found out that the target vector
> for this configuration seems to be the procfs one.
Heh, s390/s390 is Linux, so you can drop it from that list. So in
fact all the above targets, use procfs. Now the problem is that there
are a few configurations that you don't mention above that use procfs
too. It wouldn't suprise me at all if most of them would support the
gcore functionality, although OSF/1 probably needs some tweaks since
it doesn't use ELF as its executable format.
> In the rest of this discussion, I will refer to "the" target vector
> as the one at the process stratum.
>
> For Linux and FreeBSD, it was easy because the nat files were always
> creating their own target vector based on the ptrace one. The ptrace
> target vector was never pushed on the target stack.
What do you mean by "never pushed on the target stack"? There are
push_target() calls in inf-ptrace.c
> However, so far, the procfs target vector IS pushed on the target
> stack, and no modification has been needed so far. But if we are
> to continue in this direction, we will need to be able to modify
> this target vector.
> I was less and less convinced by the entire idea of using the target
> vector, at least for such a minor extension to GDB just for HP/UX.
> Perhaps a few lines of code duplication in inf-ttrace wouldn't be
> so bad.
We've made such hacks too often in the past. In fact those hacks
largely are why you're having so much trouble now.
> Except that it might be a good general cleanup for GDB. For instance,
> I really like the model used by Linux for instance, where they use
> a default target vector provided by inf-ptrace, and tailor it to
> their needs. Would it make sense to modify the procfs support such
> that it is no longer pushed itself on the stack, but instead used
> as a tailorable template?
That's exactly what needs to be done. Unfortunately npbody seems to
care enough about the targets using procfs to do the actual work.
> I'm sure we can find ways to scrub directly inside the procfs
> target vector, but this would seem rather ugly to me.
Agreed.
> The whole patch is becoming too big in my opinion to be submitted
> in one go. If we agree on continuing on this path, then I will submit
> one or more preparatory patches (such as making the procfs target vector
> a template for instance), as needed.
That'd certainly be appreciated.
Mark
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RFA] implement gcore on hp/ux
2006-01-12 20:44 ` Mark Kettenis
@ 2006-01-13 4:25 ` Michael Snyder
0 siblings, 0 replies; 6+ messages in thread
From: Michael Snyder @ 2006-01-13 4:25 UTC (permalink / raw)
To: Mark Kettenis; +Cc: brobecker, gdb-patches
Mark Kettenis wrote:
>>Date: Thu, 12 Jan 2006 09:55:29 +0400
>>From: Joel Brobecker <brobecker@adacore.com>
>
>>The whole patch is becoming too big in my opinion to be submitted
>>in one go. If we agree on continuing on this path, then I will submit
>>one or more preparatory patches (such as making the procfs target vector
>>a template for instance), as needed.
>
>
> That'd certainly be appreciated.
I'll just chime in here, since I've done a lot of work on procfs.c.
This is just FYI.
There are actually two (major) flavors of /proc, and it's been
tricky to keep procfs working smoothly for both of them. The
old-style procfs uses ioctl calls (solaris 2.5, irix...), and
the newer style uses only read and write (solaris 2.6 and beyond,
Unixware...). Be sure you test changes on at least one instance
of each.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2006-01-13 4:25 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-01-10 17:18 [RFA] implement gcore on hp/ux Joel Brobecker
2006-01-10 20:29 ` Mark Kettenis
2006-01-11 3:59 ` Joel Brobecker
2006-01-12 5:55 ` Joel Brobecker
2006-01-12 20:44 ` Mark Kettenis
2006-01-13 4:25 ` Michael Snyder
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox