This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[pushed][Cell/B.E.] Make parse_spufs_run more robust
- From: "Ulrich Weigand" <uweigand at de dot ibm dot com>
- To: gdb-patches at sourceware dot org
- Date: Thu, 27 Aug 2015 19:41:37 +0200 (CEST)
- Subject: [pushed][Cell/B.E.] Make parse_spufs_run more robust
- Authentication-results: sourceware.org; auth=none
Hello,
with recent changes to inferior handling, parse_spufs_run needs to be
more careful in assumptions it makes. In particular, this patch:
- Bails out early if the current inferior has not yet been registered
(e.g. during fork procession) to avoid assertion failures in register
cache code.
- Sets inferior_ptid to the current ptid while calling target_read_memory
to make sure the correct process is accessed if parse_spufs_run is
called early when inferior_ptid has not yet been switched by the caller.
Tested on Cell/B.E., pushed to mainline.
Bye,
Ulrich
ChangeLog:
* spu-multiarch.c (parse_spufs_run): Bail out if inferior is not
registered yet. Set inferior_ptid while calling target_read_memory.
Index: binutils-gdb/gdb/spu-multiarch.c
===================================================================
--- binutils-gdb.orig/gdb/spu-multiarch.c
+++ binutils-gdb/gdb/spu-multiarch.c
@@ -56,6 +56,7 @@ static int
parse_spufs_run (ptid_t ptid, int *fd, CORE_ADDR *addr)
{
enum bfd_endian byte_order = gdbarch_byte_order (target_gdbarch ());
+ struct cleanup *old_chain;
struct gdbarch_tdep *tdep;
struct regcache *regcache;
gdb_byte buf[4];
@@ -65,12 +66,21 @@ parse_spufs_run (ptid_t ptid, int *fd, C
if (gdbarch_bfd_arch_info (target_gdbarch ())->arch != bfd_arch_powerpc)
return 0;
+ /* If we're called too early (e.g. after fork), we cannot
+ access the inferior yet. */
+ if (find_inferior_ptid (ptid) == NULL)
+ return 0;
+
/* Get PPU-side registers. */
regcache = get_thread_arch_regcache (ptid, target_gdbarch ());
tdep = gdbarch_tdep (target_gdbarch ());
/* Fetch instruction preceding current NIP. */
- if (target_read_memory (regcache_read_pc (regcache) - 4, buf, 4) != 0)
+ old_chain = save_inferior_ptid ();
+ inferior_ptid = ptid;
+ regval = target_read_memory (regcache_read_pc (regcache) - 4, buf, 4);
+ do_cleanups (old_chain);
+ if (regval != 0)
return 0;
/* It should be a "sc" instruction. */
if (extract_unsigned_integer (buf, 4, byte_order) != INSTR_SC)
--
Dr. Ulrich Weigand
GNU/Linux compilers and toolchain
Ulrich.Weigand@de.ibm.com