gprof fix for IA-64 (and other platforms...)
Alan Modra
alan@linuxcare.com.au
Mon Mar 12 23:28:00 GMT 2001
On Mon, 12 Mar 2001, David Mosberger wrote:
> But until then, I think it's cleaner to use DEFUN in the patch
> (think of it not as new code, but re-structured code... ;-).
OK. I'm not going to press the point here. What you're saying is quite
reasonable.
> --- basic_blocks.c 2001/01/27 19:54:49 1.4
> +++ basic_blocks.c 2001/03/13 06:31:57
> @@ -227,24 +216,16 @@
> }
>
> /* Write header: */
> - bfd_put_32 (core_bfd, nblocks, (bfd_byte *) & nblocks);
> - if (fwrite (&tag, sizeof (tag), 1, ofp) != 1
> - || fwrite (&nblocks, sizeof (nblocks), 1, ofp) != 1)
> - {
> - perror (filename);
> - done (1);
> - }
> + gmon_io_write_8 (ofp, GMON_TAG_BB_COUNT);
> + gmon_io_write_32 (ofp, nblocks);
perror?
> +int
> +DEFUN (gmon_io_write_8, (ofp, valp), FILE * ofp AND u_int8_t val)
One of those pesky newfangled integer types escaped extermination.
Alan
--
Linuxcare
More information about the Binutils
mailing list