This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch 3/4] Fix hw watchpoints #2: reordered / simultaneously hit

> Date: Tue, 17 Nov 2009 16:29:12 +0100
> From: Jan Kratochvil <jan.kratochvil@redhat.com>
> Cc: Eli Zaretskii <eliz@gnu.org>, gdb-patches@sourceware.org
> 
>  0000000000000000 <get>:
>     0:  0f b6 47 04             movzbl 0x4(%rdi),%eax
> +   4:  83 e0 01                and    $0x1,%eax
>         c3                      retq   
>  
>  0000000000000010 <set>:
> -  10:  c6 47 04 01             movb   $0x1,0x4(%rdi)
> +  10:  80 4f 04 01             orb    $0x1,0x4(%rdi)

This is what I remembered, at least for x86.  A single additional
instruction can hardly cause any visible slowdown, at least not with
access patterns typical for such flags.

Any other reasons not to use bitfields?
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]