[PATCH] fix accounting of hw watchpoints on ppc

[PATCH] fix accounting of hw watchpoints on ppc

[Edjunior Barbosa Machado]

Hi,

this small patch fixes the checking of hardware watchpoints on ppc,
taking into account the different types of watches. With this change,
the following unexpected failures no longer happen:

FAIL: gdb.base/watch-read.exp: read watchpoint triggers when value doesn't change, trapping reads and writes
FAIL: gdb.base/watch-read.exp: only read watchpoint triggers when value doesn't change

Tested against ppc64/32, with no regressions found.

Ok to apply?

Thanks,
-- 
Edjunior Barbosa Machado
IBM Linux Technology Center

gdb/
2012-05-29  Edjunior Machado  <emachado@linux.vnet.ibm.com>

	* ppc-linux-nat.c (ppc_linux_can_use_hw_breakpoint): fix
	accounting of hw watchpoints on ppc.

diff --git a/gdb/ppc-linux-nat.c b/gdb/ppc-linux-nat.c
index 9bd11fd..b9e0c85 100644
--- a/gdb/ppc-linux-nat.c
+++ b/gdb/ppc-linux-nat.c
@@ -1461,7 +1461,7 @@ ppc_linux_can_use_hw_breakpoint (int type, int cnt, int ot)
   if (type == bp_hardware_watchpoint || type == bp_read_watchpoint
       || type == bp_access_watchpoint || type == bp_watchpoint)
     {
-      if (cnt > total_hw_wp)
+      if (cnt + ot > total_hw_wp)
 	return -1;
     }
   else if (type == bp_hardware_breakpoint)

Re: [PATCH] fix accounting of hw watchpoints on ppc

[Jan Kratochvil]

On Tue, 29 May 2012 20:26:59 +0200, Edjunior Barbosa Machado wrote:
> FAIL: gdb.base/watch-read.exp: read watchpoint triggers when value doesn't change, trapping reads and writes
> FAIL: gdb.base/watch-read.exp: only read watchpoint triggers when value doesn't change
> 
> Tested against ppc64/32, with no regressions found.
> 
> Ok to apply?

Yes, I find it as an improvement.  Still fixing this whole watchpointing
interface is another issue.


Thanks,
Jan

Re: [PATCH] fix accounting of hw watchpoints on ppc

[Edjunior Barbosa Machado]

On 05/30/2012 09:59 AM, Jan Kratochvil wrote:

> On Tue, 29 May 2012 20:26:59 +0200, Edjunior Barbosa Machado wrote:
>> FAIL: gdb.base/watch-read.exp: read watchpoint triggers when value doesn't change, trapping reads and writes
>> FAIL: gdb.base/watch-read.exp: only read watchpoint triggers when value doesn't change
>>
>> Tested against ppc64/32, with no regressions found.
>>
>> Ok to apply?
> 
> Yes, I find it as an improvement.  Still fixing this whole watchpointing
> interface is another issue.


Checked in:
	http://sourceware.org/ml/gdb-cvs/2012-05/msg00211.html

Thanks,
-- 
Edjunior