This is the mail archive of the gdb@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: GDB internal error: pc in psymtab, not in symtab


I suggest you read this thread:
http://sourceware.org/ml/gdb/2006-06/msg00085.html

It may be the cause of your problem, as it is the cause of this warning for me.

I simply comment out the code referred to in this post and this warning goes away for me. I can't come up with a better way to deal with it. Of course if you also have .linkonce sections in your program, you are probably hosed.

Steven J

Daniel Jacobowitz wrote:
On Wed, Oct 11, 2006 at 09:29:07PM -0700, Michael Eager wrote:
I've been bitten by this same problem:
http://sourceware.org/ml/gdb/2005-11/msg00279.html

It looks like GDB issues a warning when the ELF symbol
table contains symbols, but the DWARF data doesn't.
Assembler source with a .file will have DWARF sections,
but only a TAG_Compilation_Unit. It's legit DWARF.

Psymtabs do not come from the ELF symbol table; they come from a first pass over the DWARF information. So, if this warning triggers, that pass must have done something bogus.

"maint print psymbols" and "maint print symbols" might be useful;
the interesting question is, what psymtab thought it covered the
address, and why didn't the corresponding symtab?

Oh, one frequent cause is mis-estimating the range of addresses covered
by the psymtab.



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