[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[libunwind] Re: dwarf fixes: searching the table in eh_frame_hdr
From: |
David Mosberger |
Subject: |
[libunwind] Re: dwarf fixes: searching the table in eh_frame_hdr |
Date: |
Wed, 9 Jun 2004 17:10:16 -0700 |
>>>>> On Wed, 09 Jun 2004 16:53:48 -0700, Max Asbock <address@hidden> said:
Max> The original code allows DW_EH_PE_sdata4 for 4 byte pointer
Max> sizes and DW_EH_PE_sdata8 for 8 byte pointer sizes. This would
Max> exclude x86_64.
Ah, yes, of course. I see the problem now.
>> There must be something else that's wrong here. In fact, the new
>> code in the patch seems definitely bogus:
>> + if (!hdr->table_enc == (DW_EH_PE_datarel | DW_EH_PE_sdata4))
>> why complementing hdr->table_enc here?
Max> Yes this is bogus. It was meant to allow only (DW_EH_PE_daterel
Max> | DW_EH_PE_sdata4) since that's what the implementation in the
Max> patch allows.
OK, I think that's fine then. I think all 64-bit DWARF2 targets
currently use 32-bit tables, so limiting the entries to 32 bits should
be sufficient for a while. If and when there is a need for adding
64-bit tables, we can add the necessary routines to
Gfind_proc_info-lsb.c at that point.
What's up with the start_ip_offset -> start_ip renaming? Is this just
for shortening the names or is there a deeper reason?
--david