This is the mail archive of the gdb@sources.redhat.com 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]

fixing 'backtrace through sleep' on the cheap


In a nutshell, the problem is that gdb 6.0 does a lot worse than
gdb 5.3 when it's backtracing through functions with no debug info
and a failed prologue analysis.

I'm about to write a patch for the i386 prologue analyzer so that it
recognizes a few more instructions that are turning up in common
prologues, particularly the code for 'sleep' in glibc 2.2.93-5-rh.

I don't know whether such a patch has any chance of approval.
I'm hoping this message will flush out any disapproval
before the patch plops onto the mailing list.

Elena Z suggested this earlier today, and I think it's the best way
forward since we have a gdb 6.0 release coming shortly.

For gdb 6.1 I agree with Daniel J's suggestion to use the 'abstract
interpretation' machinery in s390-tdep.c.

Michael C


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