Core file size?

Core file size?

[linux err]

Does anyone know what determines the size of a core
dump? I have a process running out of memory (it
allocates about 3GB) - but the size of core varies
(between 2-3GB) depending on how much the process
wrote on the allocated memory.

Also, the time it takes to write the core (same size)
varies??

I briefly looked at elf_core_dump and get_user_pages()
in binfmt_elf.c. Is there any documentation on this?
Or anyone knows how it works?

TIA


 
____________________________________________________________________________________
Cheap talk?
Check out Yahoo! Messenger's low PC-to-Phone call rates.
http://voice.yahoo.com

Re: Core file size?

[Glynn Clements]

linux err wrote:

> 
> Does anyone know what determines the size of a core
> dump? I have a process running out of memory (it
> allocates about 3GB) - but the size of core varies
> (between 2-3GB) depending on how much the process
> wrote on the allocated memory.
> 
> Also, the time it takes to write the core (same size)
> varies??
> 
> I briefly looked at elf_core_dump and get_user_pages()
> in binfmt_elf.c. Is there any documentation on this?
> Or anyone knows how it works?

The relevant function seems to be maydump() (in binfmt_elf.c); this
determines whether or not a given VM area should be dumped.

Essentially, it doesn't dump data which can be obtained elsewhere,
e.g. from an executable, shared library, unmodified mmap()d file, etc.

Primarily, this prevents dumping the .text and .rodata segments of the
executable and shared libraries, which will often account for most of
a process' address space.

-- 
Glynn Clements <glynn@gclements.plus.com>

Re: Core file size?

[linux err]

This patch fixed it.

http://lkml.org/lkml/2005/6/13/204

The ELF core dump code has one use of off_t when
writing out segments.  Some of the segments may be
passed the 2GB limit of an off_t, even on a 32-bit
system, so it's important to use loff_t instead.  This
fixes a corrupted core dump in the bigcore test in
GDB's testsuite.

Signed-off-by: Daniel Jacobowitz
<dan@codesourcery.com>

Index: linux-2.6.11/fs/binfmt_elf.c
===================================================================
--- linux-2.6.11.orig/fs/binfmt_elf.c	2005-06-09
16:38:17.000000000 -0400
+++ linux-2.6.11/fs/binfmt_elf.c	2005-06-10
00:10:52.000000000 -0400
@@ -1125,7 +1125,7 @@ static int dump_write(struct
file *file,
 	return file->f_op->write(file, addr, nr,
&file->f_pos) == nr;
 }
 
-static int dump_seek(struct file *file, off_t off)
+static int dump_seek(struct file *file, loff_t off)
 {
 	if (file->f_op->llseek) {
 		if (file->f_op->llseek(file, off, 0) != off)






--- Glynn Clements <glynn@gclements.plus.com> wrote:

> 
> linux err wrote:
> 
> > 
> > Does anyone know what determines the size of a
> core
> > dump? I have a process running out of memory (it
> > allocates about 3GB) - but the size of core varies
> > (between 2-3GB) depending on how much the process
> > wrote on the allocated memory.
> > 
> > Also, the time it takes to write the core (same
> size)
> > varies??
> > 
> > I briefly looked at elf_core_dump and
> get_user_pages()
> > in binfmt_elf.c. Is there any documentation on
> this?
> > Or anyone knows how it works?
> 
> The relevant function seems to be maydump() (in
> binfmt_elf.c); this
> determines whether or not a given VM area should be
> dumped.
> 
> Essentially, it doesn't dump data which can be
> obtained elsewhere,
> e.g. from an executable, shared library, unmodified
> mmap()d file, etc.
> 
> Primarily, this prevents dumping the .text and
> .rodata segments of the
> executable and shared libraries, which will often
> account for most of
> a process' address space.
> 
> -- 
> Glynn Clements <glynn@gclements.plus.com>
> -
> To unsubscribe from this list: send the line
> "unsubscribe linux-c-programming" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at 
> http://vger.kernel.org/majordomo-info.html
> 



 
____________________________________________________________________________________
Do you Yahoo!?
Everyone is raving about the all-new Yahoo! Mail beta.
http://new.mail.yahoo.com

Re: Core file size?

[linux err]

ulimit -c is set to unlimited. (like i said core size
varies depending on how much you wrote on the memory
you allocated)



--- David Vernon <fifty.63@ladadee.com> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> linux err wrote:
> 
> > Does anyone know what determines the size of a
> core
> > dump? 
> 
> ulimit -c to set the size and ulimit -a to see the
> size.
> 
> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (GNU/Linux)
> 
>
iD8DBQFFbvDAxkzzjLpmvCwRAr95AJ9uztMsfMBuy9yPbcGZqXOYFrk62ACgycsY
> 4hWUQKuZBsthw4cFdYNSqMg=
> =LXka
> -----END PGP SIGNATURE-----
> 



 
____________________________________________________________________________________
Do you Yahoo!?
Everyone is raving about the all-new Yahoo! Mail beta.
http://new.mail.yahoo.com