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: RFC: change longjmp logic to follow exception logic


On 06/08/2012 04:21 PM, Tom Tromey wrote:

> I'd appreciate comments on this patch.
> 
> Fedora 16 has SystemTap probe points in its glibc longjmp code, and it
> ships a gdb with these probes enabled.  This means that when debugging
> gdb itself, I finally got to see how the longjmp logic worked in
> practice.
> 
> I have found it to be quite confusing to use.  The issue is that any
> longjump will cause the inferior to stop.  For example, this behavior
> means that "next"ing will often cause a stop in some random bit of code
> much further down the stack; and then I have to use multiple "finish"es
> to get somewhere close to where I started, so I can finish the "next" by
> hand.
> 
> This is basically never what I want gdb to do.


Absolutely agreed.  The patch looks good to me.

-- 
Pedro Alves


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]