[cfe-commits] [PATCH] Add configuration of CLANG_VENDOR, VENDOR_GCC_VERSION

[Ron Lieberman]

Sebastian, Thanks for adding the clarifications.

It is indeed the case that 6.0.00 is arbitrary.
The proposed change makes it easier for some company to move to 4.4.0 and
want to reflect that in the output of -dumpversion.
The capability to do so is what we are providing. 

Perhaps if I changed VENDOR_GCC_VERSION to be VENDOR_GCC_DUMPVERSION it
might be a bit more clear.

If folks are ok with that change I can proceed forward to do so, and
resubmit the two patches.

Thanks
---
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by
The Linux Foundation


-----Original Message-----
From: Sebastian Pop [mailto:spop at codeaurora.org] 
Sent: Wednesday, August 29, 2012 11:59 PM
To: Chandler Carruth
Cc: Ron Lieberman; Rafael Espíndola; llvm-commits at cs.uiuc.edu;
cfe-commits at cs.uiuc.edu
Subject: Re: [cfe-commits] [PATCH] Add configuration of CLANG_VENDOR,
VENDOR_GCC_VERSION

Chandler Carruth wrote:
> There is no GCC version 6.0.00. That doesn't even make sense as a 
> version number.
> 

Yeah, you shouldn't pay attention to the number that Ron is speaking about:
it's just an example with three random numbers x.y.z.

> I think that if you want to deeply change the versions used by clang 
> in this way for a single product you are shipping, you should do so 
> out-of-tree. Having these macros in the tree complicates the code, and 
> makes it more likely that distributions or packagers will (correctly 
> or
> incorrectly) override them. When that happens, bug reports and issues 
> with upstream Clang become harder to understand and diagnose.

I don't think this will ever be an issue.

In the case of GCC, almost every linux distro and company releasing GCC
toolchains have their own patches on top of the code of GCC. Users of these
GCC distributions are directed to submit bug reports to the linux distro or
company that released the code, and not to GCC's bugzilla. In the rear
occasions a bug is directed to GCC's bugzilla, and is identified to fail
only for a specific vendor toolchain, the bug is taken care of by the
organization who released the code.

>From this perspective it is a good thing to be able to identify these
compilers with a name and version different than that of llvm.org. In case a
bug is reported against a version that is not released by llvm.org, the bug
should be rerouted to the organization that released that code.

Sebastian
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by
The Linux Foundation