llvm dev - Dec 2013 - [LLVMdev] Reporting errors when applying fixups

For a target that hasn't implemented branch relaxation (yet),  does anyone
know what is the preferred way to report an error if a fixup cannot be applied
because, for example, the destination of a branch is out of range?
I suppose I could use asserts just like AArch64 is doing but that won't stop
the assembler of emitting a branch to an undesired location in release builds.
Does anyone see any problem in using report_fatal_error to report errors in the
target Asm backend ?

Regards,
Matheus

Matheus Almeida
MIPS processor IP
www.imgtec.com<http://www.imgtec.com/>

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20131203/0be6607c/attachment.html>

Matheus,

 

The ARM backend reports these kinds of errors using FatalError method of
MCContext. You can see some examples in ARMAsmBackend.cpp (search for "out
of range pc-relative fixup value").

 

-David

 

From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu] On
Behalf Of Matheus Almeida
Sent: Tuesday, December 03, 2013 5:37 AM
To: llvmdev at cs.uiuc.edu
Subject: [LLVMdev] Reporting errors when applying fixups

 

For a target that hasn't implemented branch relaxation (yet),  does anyone
know what is the preferred way to report an error if a fixup cannot be
applied because, for example, the destination of a branch is out of range? 

I suppose I could use asserts just like AArch64 is doing but that won't stop
the assembler of emitting a branch to an undesired location in release
builds.

Does anyone see any problem in using report_fatal_error to report errors in
the target Asm backend ? 

 

Regards,

Matheus

 

Matheus Almeida

MIPS processor IP

 <http://www.imgtec.com/> www.imgtec.com

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20131203/97693751/attachment.html>

Yup, that’s the best option available right now. There’s limited support for
adding a source location to the fixups so the diagnostic isn’t completely
useless when coming from the assembler. There’s a proposal ongoing about
improved diagnostics from the backend that Quentin is working on that will help
here when it’s completed.

-Jim

On Dec 3, 2013, at 9:51 AM, David Peixotto <dpeixott at codeaurora.org>
wrote:
> Matheus,
>  
> The ARM backend reports these kinds of errors using FatalError method of
MCContext. You can see some examples in ARMAsmBackend.cpp (search for "out
of range pc-relative fixup value").
>  
> -David
>  
> From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at
cs.uiuc.edu] On Behalf Of Matheus Almeida
> Sent: Tuesday, December 03, 2013 5:37 AM
> To: llvmdev at cs.uiuc.edu
> Subject: [LLVMdev] Reporting errors when applying fixups
>  
> For a target that hasn’t implemented branch relaxation (yet),  does anyone
know what is the preferred way to report an error if a fixup cannot be applied
because, for example, the destination of a branch is out of range?
> I suppose I could use asserts just like AArch64 is doing but that won’t
stop the assembler of emitting a branch to an undesired location in release
builds.
> Does anyone see any problem in using report_fatal_error to report errors in
the target Asm backend ?
>  
> Regards,
> Matheus
>  
> Matheus Almeida
> MIPS processor IP
> www.imgtec.com
>  
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20131203/c137cf7e/attachment.html>