llvm dev - Dec 2006 - [LLVMdev] MachineConstantPoolValue

In the ARM backend, functions (and other 32 bit constants) are placed
in a pool and loaded when needed.

A problem with this is that ".weak" directives must be printed in the
pool. This is not supported in the standard printer, so I think that I
have found the first use for MachineConstantPoolValue :-)

Creating the constant is easy, but I have two problems:

1) what are the methods getExistingMachineCPValue and
AddSelectionDAGCSEId supposed to do?

2) printing is hard because we are outside the AsmPrinter.

In my current hack I am printing from
ARMAsmPrinter::EmitMachineConstantPoolValue :-(

Since this problem might also be significant to x86 (see my previous
email), maybe we should just teach the pool printer to emit ".weak"
directives...

Best Regards,
Rafael

P.S.: rest assured that what is attached is just a demo hack and will
not be committed :-)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: llvm-pool-hack.patch
Type: text/x-patch
Size: 4362 bytes
Desc: not available
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20061206/b87d2062/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ARMAsmPrinter.h
Type: text/x-chdr
Size: 547 bytes
Desc: not available
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20061206/b87d2062/attachment.h>

On Wed, 6 Dec 2006, [UTF-8] Rafael Esp?ndola wrote:
> In the ARM backend, functions (and other 32 bit constants) are placed
> in a pool and loaded when needed.
Has this approach been replaced with your later constant pool patch?

-Chris
> A problem with this is that ".weak" directives must be printed in
the
> pool. This is not supported in the standard printer, so I think that I
> have found the first use for MachineConstantPoolValue :-)
>
> Creating the constant is easy, but I have two problems:
>
> 1) what are the methods getExistingMachineCPValue and
> AddSelectionDAGCSEId supposed to do?
>
> 2) printing is hard because we are outside the AsmPrinter.
>
> In my current hack I am printing from
> ARMAsmPrinter::EmitMachineConstantPoolValue :-(
>
> Since this problem might also be significant to x86 (see my previous
> email), maybe we should just teach the pool printer to emit
".weak"
> directives...
>
> Best Regards,
> Rafael
>
> P.S.: rest assured that what is attached is just a demo hack and will
> not be committed :-)
>
>
-Chris

-- 
http://nondot.org/sabre/
http://llvm.org/

On 12/8/06, Chris Lattner <sabre at nondot.org>
wrote:
> On Wed, 6 Dec 2006, [UTF-8] Rafael Esp?ndola wrote:
> > In the ARM backend, functions (and other 32 bit constants) are placed
> > in a pool and loaded when needed.
>
> Has this approach been replaced with your later constant pool patch?
Yes. The constant pool patch solves the problem in a much simpler way.

Best Regards,
Rafael