[PATCH][gdb] Enable some more print_one_insn selftests

[PATCH][gdb] Enable some more print_one_insn selftests

[Tom de Vries]

Hi,

In print_one_insn_test we have this cluster of skipped tests:
...
    case bfd_arch_ia64:
    case bfd_arch_mep:
    case bfd_arch_mips:
    case bfd_arch_tic6x:
    case bfd_arch_xtensa:
      return;
...

Enable some of these, and document in more detail why they're enabled or
skipped.

Likewise, document bfd_arch_or1k because it's an odd case.

Tested on x86_64-linux.

Any comments?

Thanks,
- Tom

[gdb] Enable some more print_one_insn selftests

---
 gdb/disasm-selftests.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/gdb/disasm-selftests.c b/gdb/disasm-selftests.c
index 266745d01c7..4f5667bc4e2 100644
--- a/gdb/disasm-selftests.c
+++ b/gdb/disasm-selftests.c
@@ -50,11 +50,29 @@ print_one_insn_test (struct gdbarch *gdbarch)
       len = sizeof (arm_insn);
       break;
     case bfd_arch_ia64:
+      /* We get:
+	 internal-error: gdbarch_sw_breakpoint_from_kind:
+	 Assertion `gdbarch->sw_breakpoint_from_kind != NULL' failed.  */
+      return;
     case bfd_arch_mep:
+      /* Disassembles as '*unknown*' insn, then len self-check fails.  */
+      return;
     case bfd_arch_mips:
+      if (gdbarch_bfd_arch_info (gdbarch)->mach == bfd_mach_mips16)
+	/* Disassembles insn, but len self-check fails.  */
+	return;
+      goto generic_case;
     case bfd_arch_tic6x:
+      /* Disassembles as '<undefined instruction 0x56454314>' insn, but len
+	 self-check passes, so let's allow it.  */
+      goto generic_case;
     case bfd_arch_xtensa:
+      /* Disassembles insn, but len self-check fails.  */
       return;
+    case bfd_arch_or1k:
+      /* Disassembles as '*unknown*' insn, but len self-check passes, so let's
+	 allow it.  */
+      goto generic_case;
     case bfd_arch_s390:
       /* nopr %r7 */
       static const gdb_byte s390_insn[] = {0x07, 0x07};

Re: [PATCH][gdb] Enable some more print_one_insn selftests

[Tom Tromey]

>>>>> "Tom" == Tom de Vries via Gdb-patches <gdb-patches@sourceware.org> writes:

Tom> Enable some of these, and document in more detail why they're enabled or
Tom> skipped.

Tom> Likewise, document bfd_arch_or1k because it's an odd case.

Tom> Tested on x86_64-linux.

Tom> Any comments?

Looks good.  Thanks for doing this, I appreciate the comments here.

Tom