Hi,
Thanks for your valuable pointer.
After debugging the unwinder code, we found out that the problem arises
because of optimization flags -fno-gcse and -fno-dse. After defining
LIBGCC2_UNWIND_ATTRIBUTE macro to
__attribute__((optimize("no-dse","-no-gcse"))) , things looks a bit better.
Comparing the object dumps with and without -fgcse
Code snippet with -fno-gcse
push $0x2,r0
loadd 0x50:l(r12),(r1,r0)
jal (r1,r0)
addd $0x8:s,(sp)
code snippet with -fgcse
push $0x2,r0
jal (r11,r10)
addd $0x8:s,(sp)
NOTE : r12 is PIC register
jal jumps to address specified in register pair (r1,r0)
With -fgcse optimizes PIC load instruction and jumps to address specified by
(r11,r10) register pair . But (r11,r10) has no pointer stored in it. This is
the place where my kernel gets crashed because of corrupted JAL.
How can I make sure my LOAD instruction won't get optimized away with -fgcse.
Thanks in advance,
Sumanth G
-----Original Message-----
From: Ian Lance Taylor [mailto:iant@xxxxxxxxxx]
Sent: Tuesday, November 23, 2010 10:14 PM
To: Sumanth Gundapaneni
Cc: gcc-help@xxxxxxxxxxx
Subject: Re: Exception Handling implementation with Return Address (RA) having PC right shifted by 1
Sumanth Gundapaneni <Sumanth.Gundapaneni@xxxxxxxxxxxxxxx> writes:
> The problem arises with throw and catch code in c++ EH code. During the runtime unwinding process of stack, RA register value which is a right shifted value of PC (because of hardware implementation) is interpreted by stack unwinder as is. So the call backs across functions were pointed to wrong locations.
Hmmm, I see.
> How can I adjust the RA (return address) so that my throw, catch EH code works fine with my compiler.
To be clear, I don't think you want to adjust the RA. I think you want
to adjust the DWARF encoding of the RA, so that the DWARF code knows
what to do with the value. As I said earlier, the usual way to tell the
DWARF code to do something unusual is to build a REG_CFA_DEF_CFA note
and attach it to some relevant insn. This is an unusual case which I've
never seen before, and I don't know exactly how to deal with it. But I
think REG_CFA_DEF_CFA is the right area to look.
Ian
