Re: [PATCH 0/2] Global breakpoints, introduction

[Kevin Pouget <kevin.pouget@gmail.com>]

Hello,

the idea is very interesting, I'll certainly have to use it soon!

>  fprintf_unfiltered (gdb_stdlog, "target_to_define_global_breakpoint (0x%x, %s, %s, %d) = %d\n",
>		      (int) abfd, paddress (target_gdbarch, addr),
>		      (uname ? uname : "<NULL>"), flags, rslt);

target.c:4200: error: cast from pointer to integer of different size
[-Werror=pointer-to-int-cast]


and in the README, you mention the file "/proc/drivers/breakpoints"
whereas it's certainly /proc/driver/breakpoint


cordially,

Kevin

On Mon, Jun 13, 2011 at 7:03 PM, Stan Shebs <stanshebs@earthlink.net> wrote:
>
> On 6/13/11 9:39 AM, Frank Ch. Eigler wrote:
>>
>> Hi, Stan -
>>
>> stanshebs wrote:
>>
>>> This pair of patches plus Linux kernel module adds global breakpoints
>>> to GDB.  [...]
>>
>> Interesting approach.  Have you been planning to post it to LKML for
>> review/consideration?
>>
>
> Yes, I've just been waiting for all the debugger-haters there to retire and/or die first. :-)
>
>> You may want to rebase your code to the LKML uprobes implementation
>> already under review, so you don't have to duplicate the rather
>> intricate logic required to safely manage breakpoints in userspace by
>> the kernel.
>
> That's a good idea!  I haven't updated myself on uprobes state in the past several months, but you're right, it would be good to have a single version of that.
>
> Stan
>