Hi all,
There's a small problem that's brewing with the exception handling
stuff that some of us have been working on lately. The problem is that
different architectures are going to require different data format
encodings for different exception handling bits. For example, one
architecture may want PCREL for an entry, while another may want PCREL
+ SDATA4. There is a problem with the comment line for such
information in the ASM file. We currently only have a mechanism in
place to emit a string as a comment followed by EOL.
The attached patch is an attempt to follow what GCC does. And by
attempt, I mean that I whole-sale copied the code over from one of
their functions, which returns the string of the format encoding in
question.
My RFC is two-fold:
1. Is this okay, in both a legal and moral sense?
2. Is there a better way to do this?
Comments?
Please note that the patch isn't wonderful. It's more of a proof-of-
concept at this point...
-bw
-------------- next part --------------
A non-text attachment was scrubbed...
Name: designated-inits.diff
Type: application/octet-stream
Size: 15160 bytes
Desc: not available
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20090825/37a97ef1/attachment.obj>
-------------- next part --------------