[Crash-utility] [PATCH] crash: dis: introduce count in reverse and forward mode

Aaron Tomlin atomlin at redhat.com
Tue Jun 18 21:51:47 UTC 2019


On Tue 2019-06-18 16:27 -0400, Dave Anderson wrote:
[ ... ]

Hi Dave,

> This seems to have uncovered a problem with the gdb-7.6.patch scheme,
> where if the top-level Makefile sees that it has been modified, it re-applies
> it using -N and -r- here:
> 
>   patch -N -p0 -r- --fuzz=0 < ${GDB}.patch
> 
> It's currently attempting to re-apply a previously-applied printcmd.c patch 
> (i.e., not recognizing that the patch had been applied), and it is failing
> that attempt, and so the build process continues on.  Theoretically
> that should still work, but it's somehow affecting the application of your
> patch.  I'm not exactly sure why?  If you build with "make > log 2>&1", you'll see
> what I mean w/respect to the subsequent re-application of the updated gdb-7.6.patch.
> Presumably it has something to do with patching the same area of a file more
> than once.  

Intereseting, for some reason, I don't see the same issue after a
'git clean -x -d -f && git checkout -- Makefile'.
The following is an excerpt from 'make warn':

patching file gdb-7.6/gdb/printcmd.c
if [ "x86_64" = "ppc64le" ] && [ -f gdb-7.6-ppc64le-support.patch ]; then \
	patch -d gdb-7.6 -p1 -F0 < gdb-7.6-ppc64le-support.patch ; \
fi
if [ "x86_64" = "x86_64" ] && [ "X86_64" = "PPC64" ] && [ -f gdb-7.6-ppc64le-support.patch ]; then \
	patch -d gdb-7.6 -p1 -F0 < gdb-7.6-ppc64le-support.patch ; \
fi
if [ -f /usr/include/proc_service.h ]; then \
	grep 'extern ps_err_e ps_get_thread_area (struct' /usr/include/proc_service.h; \
	if [ $? -eq 0 ]; then \
		patch -p0 < gdb-7.6-proc_service.h.patch; \
	fi; \
fi
extern ps_err_e ps_get_thread_area (struct ps_prochandle *,
patching file gdb-7.6/gdb/gdb_proc_service.h


Kind regards,

-- 
Aaron Tomlin




More information about the Crash-utility mailing list