[rfc] [13/17] Get rid of current_gdbarch in win32-nat.c

[rfc] [13/17] Get rid of current_gdbarch in win32-nat.c

[Markus Deuling]

[-- Attachment #1: Type: text/plain, Size: 434 bytes --]

Hi,

this patch gets rid of some of the current_gdbarch's in win32-nat.c. Maybe someone can test this? I have no idea how to :-)
Thanks a lot. Is this ok to commit?

ChangeLog: 


	* win32-nat.c (do_win32_fetch_inferior_registers, win32_resume)
	(do_win32_store_inferior_registers): Use get_regcache_arch to get at
	the current architecture by regcache.



-- 
Markus Deuling
GNU Toolchain for Linux on Cell BE
deuling@de.ibm.com





[-- Attachment #2: diff-win32-nat --]
[-- Type: text/plain, Size: 1190 bytes --]

diff -urpN src/gdb/win32-nat.c dev/gdb/win32-nat.c
--- src/gdb/win32-nat.c	2007-09-04 03:12:18.000000000 +0200
+++ dev/gdb/win32-nat.c	2007-10-11 10:06:39.000000000 +0200
@@ -397,7 +397,7 @@ do_win32_fetch_inferior_registers (struc
     regcache_raw_supply (regcache, r, context_offset);
   else
     {
-      for (r = 0; r < gdbarch_num_regs (current_gdbarch); r++)
+      for (r = 0; r < gdbarch_num_regs (get_regcache_arch (regcache)); r++)
 	do_win32_fetch_inferior_registers (regcache, r);
     }
 
@@ -424,7 +424,7 @@ do_win32_store_inferior_registers (const
 			  ((char *) &current_thread->context) + mappings[r]);
   else
     {
-      for (r = 0; r < gdbarch_num_regs (current_gdbarch); r++)
+      for (r = 0; r < gdbarch_num_regs (get_regcache_arch (regcache)); r++)
 	do_win32_store_inferior_registers (regcache, r);
     }
 }
@@ -1211,7 +1211,8 @@ win32_resume (ptid_t ptid, int step, enu
 	{
 	  /* Single step by setting t bit */
 	  win32_fetch_inferior_registers (get_current_regcache (),
-					  gdbarch_ps_regnum (current_gdbarch));
+	  				  gdbarch_ps_regnum (get_regcache_arch
+					    (get_current_regcache ())));
 	  th->context.EFlags |= FLAG_TRACE_BIT;
 	}
 

Re: [rfc] [13/17] Get rid of current_gdbarch in win32-nat.c

[Christopher Faylor]

On Thu, Oct 11, 2007 at 10:54:14AM +0200, Markus Deuling wrote:
> this patch gets rid of some of the current_gdbarch's in win32-nat.c. Maybe 
> someone can test this? I have no idea how to :-)
> Thanks a lot. Is this ok to commit?
>
> ChangeLog: 
>
> 	* win32-nat.c (do_win32_fetch_inferior_registers, win32_resume)
> 	(do_win32_store_inferior_registers): Use get_regcache_arch to get at
> 	the current architecture by regcache.

I'm sorry.  I'm a little confused.  If you're offering a patch why can't
you test it?  Wouldn't that be considered a prerequisite for a patch?

cgf

Re: [rfc] [13/17] Get rid of current_gdbarch in win32-nat.c

[Markus Deuling]

Hi Christopher,

Christopher Faylor schrieb:
> On Thu, Oct 11, 2007 at 10:54:14AM +0200, Markus Deuling wrote:
>> this patch gets rid of some of the current_gdbarch's in win32-nat.c. Maybe 
>> someone can test this? I have no idea how to :-)
>> Thanks a lot. Is this ok to commit?
>>
>> ChangeLog: 
>>
>> 	* win32-nat.c (do_win32_fetch_inferior_registers, win32_resume)
>> 	(do_win32_store_inferior_registers): Use get_regcache_arch to get at
>> 	the current architecture by regcache.
> 
> I'm sorry.  I'm a little confused.  If you're offering a patch why can't
> you test it?  Wouldn't that be considered a prerequisite for a patch?
> 
> cgf
> 

you're right. One should test a patch before posting it ;-) But this patch is a mechanical change. Thats the reason
I ask a responsible maintainer to have a look at it instead of setting up an environment to test GDB on Windows (cygwin) myself.

Thank you for your objection and sry for the late response. I've been on vacation.

-- 
Markus Deuling
GNU Toolchain for Linux on Cell BE
deuling@de.ibm.com