[RFA]: patch to ia64.inc in gdb.asm testsuite

[RFA]: patch to ia64.inc in gdb.asm testsuite

[Jeff Johnston]

[-- Attachment #1: Type: text/plain, Size: 686 bytes --]

The attached patch is needed to get the gdb.asm testsuite running 
successfully for ia64.  A recent patch changed the link step to use the 
target linker rather than gcc.  The old method was defaulting dynamic 
loading.  The new code causes a static executable to be created with the 
first insn being that of gdbasm_startup.  Previously, the first insn was 
just a mov, but this should be an alloc insn.  Without this change, a 
SIGILL is caused for the first insn and the tests all fail.

Ok to commit?

-- Jeff J.


2004-02-04  Jeff Johnston  <jjohnstn@redhat.com>

        * gdb.asm/ia64.inc (gdbasm_startup): Make sure first instruction
        executed is an alloc instruction.



[-- Attachment #2: ia64asm.patch --]
[-- Type: text/plain, Size: 387 bytes --]

Index: ia64.inc
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.asm/ia64.inc,v
retrieving revision 1.1
diff -u -r1.1 ia64.inc
--- ia64.inc	25 Aug 2003 23:43:40 -0000	1.1
+++ ia64.inc	5 Feb 2004 00:34:57 -0000
@@ -36,6 +36,7 @@
 
 	comment "crt0 startup"
 	.macro gdbasm_startup
+	gdbasm_enter
 	mov r32=r0
 	nop.i 0
 	nop.i 0

Re: [RFA]: patch to ia64.inc in gdb.asm testsuite

[Andrew Cagney]

> The attached patch is needed to get the gdb.asm testsuite running successfully for ia64.  A recent patch changed the link step to use the target linker rather than gcc.  The old method was defaulting dynamic loading.  The new code causes a static executable to be created with the first insn being that of gdbasm_startup.  Previously, the first insn was just a mov, but this should be an alloc insn.  Without this change, a SIGILL is caused for the first insn and the tests all fail.

If it works for you :-)  I'm sure there's going to be more of this fallout.

Andrew


> 2004-02-04  Jeff Johnston  <jjohnstn@redhat.com>
> 
>        * gdb.asm/ia64.inc (gdbasm_startup): Make sure first instruction
>        executed is an alloc instruction.

Re: [RFA]: patch to ia64.inc in gdb.asm testsuite

[Jeff Johnston]

Patch checked in.  Thanks.

-- Jeff J.

Andrew Cagney wrote:
>> The attached patch is needed to get the gdb.asm testsuite running 
>> successfully for ia64.  A recent patch changed the link step to use 
>> the target linker rather than gcc.  The old method was defaulting 
>> dynamic loading.  The new code causes a static executable to be 
>> created with the first insn being that of gdbasm_startup.  Previously, 
>> the first insn was just a mov, but this should be an alloc insn.  
>> Without this change, a SIGILL is caused for the first insn and the 
>> tests all fail.
> 
> 
> If it works for you :-)  I'm sure there's going to be more of this fallout.
> 
> Andrew
> 
> 
>> 2004-02-04  Jeff Johnston  <jjohnstn@redhat.com>
>>
>>        * gdb.asm/ia64.inc (gdbasm_startup): Make sure first instruction
>>        executed is an alloc instruction.
> 
> 
> 
>