[dpdk-dev] mk: Work around Debian/Ubuntu-specific 'gcc -dumpversion' output

  Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation
on some versions of Debian and Ubuntu where gcc has been modified
to only emit MAJOR.MINOR part of the version from 'gcc -dumpversion'.
Drop the micro-version from gcc version comparisons to work around
this, it wasn't being used for anything anyway.

Signed-off-by: Panu Matilainen <pmatilai@redhat.com>
---
 lib/librte_pmd_fm10k/Makefile            |  2 +-
 lib/librte_pmd_i40e/Makefile             |  2 +-
 lib/librte_pmd_ixgbe/Makefile            |  6 +++---
 lib/librte_pmd_vmxnet3/Makefile          |  2 +-
 mk/toolchain/gcc/rte.toolchain-compat.mk | 16 ++++++++--------
 5 files changed, 14 insertions(+), 14 deletions(-)
  

Panu,

ack (and works with my debian gcc).
  

Hi Panu, David,

On 02/24/2015 11:59 AM, David Marchand wrote:
>> >Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation
>> >on some versions of Debian and Ubuntu where gcc has been modified
>> >to only emit MAJOR.MINOR part of the version from 'gcc -dumpversion'.
>> >Drop the micro-version from gcc version comparisons to work around
>> >this, it wasn't being used for anything anyway.
>> >
>> >Signed-off-by: Panu Matilainen<pmatilai@redhat.com>
>
> ack (and works with my debian gcc).
>

One question: will it work with a version like with 2 digits
for major or minor like 4.10.0 or 10.0.0?
(or maybe it cannot happen with gcc naming conventions?)

Thanks,
Olivier
  

On 02/24/2015 01:14 PM, Olivier MATZ wrote:
> Hi Panu, David,
>
> On 02/24/2015 11:59 AM, David Marchand wrote:
>>> >Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation
>>> >on some versions of Debian and Ubuntu where gcc has been modified
>>> >to only emit MAJOR.MINOR part of the version from 'gcc -dumpversion'.
>>> >Drop the micro-version from gcc version comparisons to work around
>>> >this, it wasn't being used for anything anyway.
>>> >
>>> >Signed-off-by: Panu Matilainen<pmatilai@redhat.com>
>>
>> ack (and works with my debian gcc).
>>
>
> One question: will it work with a version like with 2 digits
> for major or minor like 4.10.0 or 10.0.0?
> (or maybe it cannot happen with gcc naming conventions?)

The original change mentions this:

     This [...] scheme would run into trouble if gcc ever went to
     two-digit version segments, but that hasn't happened in the last 10+
     years so it seems like a safe assumption.

...but that's not entirely accurate, 4.10.0 would be a problem but 
10.0.0 not.

	- Panu -
  

2015-02-24 11:59, David Marchand:
> Panu,
> 
> ack (and works with my debian gcc).
> 
> > Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation
> > on some versions of Debian and Ubuntu where gcc has been modified
> > to only emit MAJOR.MINOR part of the version from 'gcc -dumpversion'.
> > Drop the micro-version from gcc version comparisons to work around
> > this, it wasn't being used for anything anyway.
> >
> > Signed-off-by: Panu Matilainen <pmatilai@redhat.com>

Applied, thanks
  

> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Panu Matilainen
> Sent: Tuesday, February 24, 2015 10:47 AM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH] mk: Work around Debian/Ubuntu-specific 'gcc -
> dumpversion' output
> 
> Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation on some
> versions of Debian and Ubuntu where gcc has been modified to only emit
> MAJOR.MINOR part of the version from 'gcc -dumpversion'.
> Drop the micro-version from gcc version comparisons to work around this,
> it wasn't being used for anything anyway.
> 
> -ifeq ($(shell test $(GCC_VERSION) -ge 440 && echo 1), 1)
> +ifeq ($(shell test $(GCC_VERSION) -ge 44 && echo 1), 1)

Hi,

I see this has been applied already but to me this version is less clear than the previous patch version.

And are we sure we won't need the micro-version. If we do, to account for something broken in a point release of gcc, we will have to redo this patch again.

John
  

On 02/24/2015 03:34 PM, Mcnamara, John wrote:
>> -----Original Message-----
>> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Panu Matilainen
>> Sent: Tuesday, February 24, 2015 10:47 AM
>> To: dev@dpdk.org
>> Subject: [dpdk-dev] [PATCH] mk: Work around Debian/Ubuntu-specific 'gcc -
>> dumpversion' output
>>
>> Commit 71f0ab1849b4fc3ca928deb566df12ca725ed150 broke compilation on some
>> versions of Debian and Ubuntu where gcc has been modified to only emit
>> MAJOR.MINOR part of the version from 'gcc -dumpversion'.
>> Drop the micro-version from gcc version comparisons to work around this,
>> it wasn't being used for anything anyway.
>>
>> -ifeq ($(shell test $(GCC_VERSION) -ge 440 && echo 1), 1)
>> +ifeq ($(shell test $(GCC_VERSION) -ge 44 && echo 1), 1)
>
> Hi,
>
> I see this has been applied already but to me this version is less clear than the previous patch version.
>
> And are we sure we won't need the micro-version. If we do, to account for something broken in a point release of gcc, we will have to redo this patch again.

Sure? I'm not enough of a fool to say yes to a question like that. I 
just know it hasn't been needed before, and it seems unlikely it will be 
needed later on either, gcc micro-version updates tend to be conservative.

My only interest here was to fix the version comparisons in a way that 
makes gcc 5 naturally supported, and I dont really give a damn whether 
there's a micro-version involved or not :) If you feel strongly about 
it, send a patch. I'm not going to object.

	- Panu -