Re: [PATCH 3/3] Remove HP-UX references fom testsuite

[Pedro Alves <palves@redhat.com>]

I looked this one over too.  A few minor comments below, but
otherwise looks good to me.  Thanks for doing this!

On 12/19/2015 11:30 PM, Simon Marchi wrote:

> 	* gdb.multi/bkpt-multi-exec.ex: Likewise.p

Typo: "ex: Likewise.p" -> "exp: Likewise."


> +gdb_test_multiple "catch vfork" "$name" {
> +    -re "Catchpoint \[0-9\]* .vfork..*$gdb_prompt $" {
> +	pass $name
> +    }
> +    -re "Catch of vfork not yet implemented.*$gdb_prompt $" {

This case can be removed.  GDB doesn't ever output this.

> +	pass $name
>      }
>  }
>



> --- a/gdb/testsuite/gdb.base/display.exp
> +++ b/gdb/testsuite/gdb.base/display.exp
> @@ -194,15 +194,9 @@ gdb_test "print/r j" " = 0x0\[\\r\\n\]+" "debug test output 2a"
>  gdb_test "print j"   " = 0\[\\r\\n\]+"   "debug test output 3"
>
>  # x/0 j doesn't produce any output and terminates PA64 process when testing
> -if [istarget "hppa2.0w-hp-hpux11*"] {
> -    xfail "'x/0 j' terminates PA64 process - skipped test point"
> -} else {
> -    gdb_test_no_output "x/0 j"
> -}
> -if [istarget "hppa*-hp-hpux*"] {
> -    # on HP-UX you could access the first page without getting an error
> -    gdb_test "x/rx j" ".*(Cannot access|Error accessing) memory.*|.*0xa:\[ \t\]*\[0-9\]+.*"
> -}
> +gdb_test_no_output "x/0 j"
> +
> +gdb_test "x/rx j" ".*(Cannot access|Error accessing) memory.*|.*0xa:\[ \t\]*\[0-9\]+.*"

This last one used to only run on hpux.  It seems like it'll now fail
on is_address_zero_readable targets -- the non-error alternative assumes
the variable's value is "0xa" ?

You could guard it with "if [is_address_zero_readable]", but, that
proc does "x 0" internally, so it sounds like it'd be pointless?

>  gdb_test "print/0 j" ".*Item count other than 1 is meaningless.*" "print/0 j"
>  gdb_test "print/s sum" " = 1000" "ignored s"









> --- a/gdb/testsuite/gdb.dwarf2/pr10770.c
> +++ b/gdb/testsuite/gdb.dwarf2/pr10770.c
> @@ -4,7 +4,6 @@
>  /* HP-UX libunwind.so doesn't provide _UA_END_OF_STACK */
>  /* { dg-do run } */
>  /* { dg-options "-fexceptions" } */
> -/* { dg-skip-if "" { "ia64-*-hpux11.*" }  { "*" } { "" } } */
>  /* Verify DW_OP_* handling in the unwinder.  */
>  
>  #include <unwind.h>

You should remove the HP-UX reference at the top too then.
But, GDB doesn't use the dg-* markers.  Note the comment at the top:

/* This file comes from GCC.  If you are tempted to change it,
   consider also changing the copy there.  */

I'd either leave it be, or, remove the whole dg-* comment block.

Comparing our copy to the gcc copy, that's the only bit that diverged so far:

$ diff -up ~/gdb/mygit/src/gdb/testsuite/gdb.dwarf2/pr10770.c  src/gcc/testsuite/gcc.dg/cleanup-13.c
--- /home/pedro/gdb/mygit/src/gdb/testsuite/gdb.dwarf2/pr10770.c        2015-05-08 17:35:43.878768249 +0100
+++ src/gcc/testsuite/gcc.dg/cleanup-13.c       2015-08-22 16:16:35.514064275 +0100
@@ -1,10 +1,8 @@
-/* This file comes from GCC.  If you are tempted to change it,
-   consider also changing the copy there.  */
-
 /* HP-UX libunwind.so doesn't provide _UA_END_OF_STACK */
 /* { dg-do run } */
 /* { dg-options "-fexceptions" } */
 /* { dg-skip-if "" { "ia64-*-hpux11.*" }  { "*" } { "" } } */
+/* { dg-skip-if "" { ! nonlocal_goto } { "*" } { "" } } */
 /* Verify DW_OP_* handling in the unwinder.  */

 #include <unwind.h>

Thanks,
Pedro Alves