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,7.3] Fix JIT clang-lli gdb-7.3 regression Re: [gdb-7.3] Error in gdb-llvm integration: Unable to read JIT descriptor from remote memory!


On Thu, 07 Jul 2011 00:43:50 +0200, Paul Pluzhnikov wrote:
> This then leaves the "unload/reload" case not handled.

What I was trying to exploit with the countercases is that breakpoint_re_set
is designed so that if anything in the inferior changes everything gets
re-set.

Since the former Paul's change 739571cc3151651f49f7171cfd98275d983bfaaa some
information is kept even across breakpoint_re_set which I find against the
design of breakpoint_re_set.  It now keeps information about inferior even if
some unspecific inferior change has happened.

The problem is breakpoint_re_set is no longer performance sufficient.  This is
a part of the large task Gary Benson is working on, to make inferior solib
notififications local and specific to each solib.  But it needs to pass more
information what has changed than the global breakpoint_re_set event.  I do
not think that workarounding breakpoint_re_set problems at its callees is
right.  But I do not have any further counter testcases.


Thanks,
Jan


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