[gentoo-user] GCC-4.3.2

[gentoo-user] GCC-4.3.2

[alain.didierjean]

Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
performed  "emerge system" ?
What gives ? Any problem ? Is it worth it right now ? Please tell...

--
 ~adj~

Re: [gentoo-user] GCC-4.3.2

[Thomas Kahle]

[-- Attachment #1: Type: text/plain, Size: 1171 bytes --]

alain.didierjean@free.fr wrote:
> Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
> performed  "emerge system" ?
> What gives ? Any problem ? Is it worth it right now ? Please tell...
> 
> --
>  ~adj~
> 

Yes, I did it at the beginning of the year, it was more or less
painfree. There were a couple of compile failures with stable packages
when rebuilding the world, but typically these are already fixed and in
bugzilla, so you can go ahead by installing the ~x86 package. (See
http://bugs.gentoo.org/show_bug.cgi?id=245160 for a tracker of the packages)
I did not notice any difference in using the system everyday, I did it
only because I can use -march=core2 now and I wanted to work with the
newer gcc when programming other things.
You should follow the upgrade guide for maximal pleasure:
http://www.gentoo.org/doc/en/gcc-upgrading.xml

hope to help
Thom


-- 
Thomas Kahle

The fundamental theorem of algebra is open source. Like any other
mathematical theorem it can be applied free of charge and everybody
has access to its proof and can convince himself how it works. Why
should software be any different?


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 261 bytes --]

Re: [gentoo-user] GCC-4.3.2

[Masood Ahmed]

[-- Attachment #1: Type: text/plain, Size: 481 bytes --]

alain.didierjean@free.fr writes:

> Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
> performed  "emerge system" ?

I did upgrade to 4.3 gcc series from 4.1. There was no compile error's
(I'm using ~x86 branch). Both emerge -e system and emerge -e world done
using new gcc.

> What gives ? Any problem ? Is it worth it right now ? Please tell...

I like the -march=native and -mtune=native options. It's the new thing
in gcc 4.3 series.

Regards,
Masood Ahmed

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: [gentoo-user] GCC-4.3.2

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 278 bytes --]

On Fri, 03 Apr 2009 17:47:25 +0530, Masood Ahmed wrote:

> I like the -march=native and -mtune=native options.

As long as you don't use distcc with differing hardware.


-- 
Neil Bothwick

"Ubuntu" is an ancient African word, meaning "I can't configure
Slackware".

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] GCC-4.3.2

[Jerry McBride]

On Friday 03 April 2009 05:15:21 pm Neil Bothwick wrote:
> On Fri, 03 Apr 2009 17:47:25 +0530, Masood Ahmed wrote:
> > I like the -march=native and -mtune=native options.
>
> As long as you don't use distcc with differing hardware.

I did that by accident once and trust me, it's a unforgettable lesson.

Jerry.





-- 

*****************************************************************************
                                       
                             From the desk of:
                             Jerome D. McBride
                                       
    17:27:39 up 3 days, 22:12,  4 users,  load average: 0.08, 0.02, 0.01
 
*****************************************************************************

Re: [gentoo-user] GCC-4.3.2

[Mike Kazantsev]

[-- Attachment #1: Type: text/plain, Size: 327 bytes --]

On Fri, 03 Apr 2009 17:47:25 +0530
Masood Ahmed <masood.ahmed09@gmail.com> wrote:

> alain.didierjean@free.fr writes:
> 
> I like the -march=native and -mtune=native options. It's the new thing
> in gcc 4.3 series.

-mtune=native can be dropped if -march=native is there already.

-- 
Mike Kazantsev // fraggod.net

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] GCC-4.3.2

[Graham Murray]

Mike Kazantsev <mike_kazantsev@fraggod.net> writes:

> -mtune=native can be dropped if -march=native is there already.

It is still worthwhile keeping it in CFLAGS as some packages remove or
replace the '-march' flag. 

Re: [gentoo-user] GCC-4.3.2

[alain.didierjean]

Selon alain.didierjean@free.fr:

> Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
> performed  "emerge system" ?
> What gives ? Any problem ? Is it worth it right now ? Please tell...

Thanks for the fast answers. It seems there's no problems...

--
~adj~

Re: [gentoo-user] GCC-4.3.2

[Christopher Walters]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

alain.didierjean@free.fr wrote:
> Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
> performed  "emerge system" ?
> What gives ? Any problem ? Is it worth it right now ? Please tell...
> 
> --
>  ~adj~

I am afraid I can't really answer this question, since I am using GCC-4.3.3.  I
can say that I have only had a couple of merge failures with that, and these
were solved by either merging something else first, or by removing an
unnecessary USE flag - something like that.

Regards,
Chris

PS: I got more such problems with 4.1.2
-----BEGIN PGP SIGNATURE-----

iQIcBAEBCgAGBQJJ1wZIAAoJEIAhA8M9p9DARgIP/j7sUVLyj1I4sjst1Reyg91O
blnC2l+vD0LVsg/P0ZQefKdPC9QKp10JSCeio7qztwx/ZGfLahH1AbXhQS45/7aM
qyJJa1qTin6wsksONy9gksnVP1gGWERv8C4BYldbl1cMGiEQfSifUmYrFY2FtOKl
o5gpp67y+fdbzwjAwSqdlqMAyHmFq9MfY+O86S9r57rrhhfcZnXdtf2ufgpNi+Kn
02bC1HbXwjYXLAtFEZCG5ajl+6+gl5JhDrnj5V1+Hz3Ouh53OryUH1OXyzR0rvWq
pR2WVraZxxKaxGzRT99a1+AVykkbY7RJI37l9WNg/ZSDdtDrl6twe7DHu6HAHzof
IHItzF/CjT7ShuTot2zfURJhSraOr3K44ZLVaOTKMGOux1F5yraYujoA0cuf21S9
tV4yAYt9oc5vWiHTfAUGz7X+tduVPola3G7RdWibsnbUMZFcgltvGKCNU0XLMXSs
wpTCYcmuQ05ui0BesScUMrtmZAjGzbVAXkRUzEO7W2r+s7E74rkbSFSsLevhT5X9
TYATTXeUsXSMgbDnAb9BLw+ZPVua9pCnJYkd3qALUSbKbn0COtZqzJXtMCRZghat
3FeX7sr9mWL48MrjGCthIvx5oDq4i+SrPhgWqg2MwSFEb6lgWp8LY/OoF2UU4oBO
zsIKSxYJeC9tz0Yq+i9R
=ITyG
-----END PGP SIGNATURE-----

Re: [gentoo-user] GCC-4.3.2

[Joseph]

On 04/03/09 10:34, alain.didierjean@free.fr wrote:
>Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
>performed  "emerge system" ?
>What gives ? Any problem ? Is it worth it right now ? Please tell...
>
>--
> ~adj~

I just notice that there is a problem with CFLAGS -march=native on AMD64
I recompile my system using this new flag and have a lot of problems. 
an examples new xorg-server-1.5... would not compile, I switch back to my previous settings:
-march=athlon64 and it compiled on first pass.
So, now I'm recompiling the system with my old flag: -march=athlon64 :-/

-- 
Joseph

[gentoo-user] Re: GCC-4.3.2

[Nikos Chantziaras]

Joseph wrote:
> On 04/03/09 10:34, alain.didierjean@free.fr wrote:
>> Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
>> performed  "emerge system" ?
>> What gives ? Any problem ? Is it worth it right now ? Please tell...
>>
>> -- 
>> ~adj~
> 
> I just notice that there is a problem with CFLAGS -march=native on AMD64
> I recompile my system using this new flag and have a lot of problems. an 
> examples new xorg-server-1.5... would not compile, I switch back to my 
> previous settings:
> -march=athlon64 and it compiled on first pass.
> So, now I'm recompiling the system with my old flag: -march=athlon64 :-/

-march=native is only useful if you don't know what to use.  I guess 
most Gentoo users do know what to use.  Why should I use "native" if I 
know that my CPU is "athlon64" :P

Re: [gentoo-user] Re: GCC-4.3.2

[Joseph]

On 04/06/09 11:56, Nikos Chantziaras wrote:

[snip]
>>> ~adj~
>> I just notice that there is a problem with CFLAGS -march=native on AMD64
>> I recompile my system using this new flag and have a lot of problems. an 
>> examples new xorg-server-1.5... would not compile, I switch back to my 
>> previous settings:
>> -march=athlon64 and it compiled on first pass.
>> So, now I'm recompiling the system with my old flag: -march=athlon64 :-/
>
> -march=native is only useful if you don't know what to use.  I guess most 
> Gentoo users do know what to use.  Why should I use "native" if I know that 
> my CPU is "athlon64" :P

Well, according to gentoo recommendations it is not so:
http://en.gentoo-wiki.com/wiki/Safe_Cflags
"...GCC 4.2 introduces a new -march option, -march=native, which automatically detects the features your CPU supports and sets the options appropriately. 
If you have an Intel or AMD CPU and are using >=sys-devel/gcc-4.2.3, using -march=native is recommended...."

So, I suggest you make a correction to gentoo-wiki, as I suspect I'll will not be the only one who get confused by it :-/

-- 
Joseph

Re: [gentoo-user] Re: GCC-4.3.2

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 343 bytes --]

On Mon, 6 Apr 2009 10:35:56 -0600, Joseph wrote:

> Well, according to gentoo recommendations it is not so:
> http://en.gentoo-wiki.com/wiki/Safe_Cflags

Information on gentoo-wiki.com is not a Gentoo recomendation. Look at
docs.gentoo.org for those.
 

-- 
Neil Bothwick

Anything worth fighting for is worth fighting dirty for.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Re: GCC-4.3.2

[Joseph]

On 04/06/09 21:07, Neil Bothwick wrote:
>On Mon, 6 Apr 2009 10:35:56 -0600, Joseph wrote:
>
>> Well, according to gentoo recommendations it is not so:
>> http://en.gentoo-wiki.com/wiki/Safe_Cflags
>
>Information on gentoo-wiki.com is not a Gentoo recomendation. Look at
>docs.gentoo.org for those.
> 
>
>-- 
>Neil Bothwick

If you have look for it more carefully you would have found similar recommendation on the official gentoo pages:
http://www.gentoo.org/doc/en/distcc.xml
I think the above page qualify as an official doc. isn't it? 

"...-march=native

Starting with GCC 4.3.0, the compiler supports the -march=native switch which turns on CPU autodetection and optimizations that are worth being enabled on 
the processor the GCC is running at..."

So reading it, it seems to me the "-march=native" is recommended as well.

-- 
Joseph

Re: [gentoo-user] Re: GCC-4.3.2

[Paul Hartman]

On Mon, Apr 6, 2009 at 3:28 PM, Joseph <syscon780@gmail.com> wrote:
> On 04/06/09 21:07, Neil Bothwick wrote:
>>
>> On Mon, 6 Apr 2009 10:35:56 -0600, Joseph wrote:
>>
>>> Well, according to gentoo recommendations it is not so:
>>> http://en.gentoo-wiki.com/wiki/Safe_Cflags
>>
>> Information on gentoo-wiki.com is not a Gentoo recomendation. Look at
>> docs.gentoo.org for those.
>>
>>
>> --
>> Neil Bothwick
>
> If you have look for it more carefully you would have found similar
> recommendation on the official gentoo pages:
> http://www.gentoo.org/doc/en/distcc.xml
> I think the above page qualify as an official doc. isn't it?
> "...-march=native
>
> Starting with GCC 4.3.0, the compiler supports the -march=native switch
> which turns on CPU autodetection and optimizations that are worth being
> enabled on the processor the GCC is running at..."
>
> So reading it, it seems to me the "-march=native" is recommended as well.

That looks like a description of what -march=native is, not a
recommendation. The recommendation is in the following sentences when
they warn NOT to use it when using distcc in certain configurations
(same as the wiki page).

Re: [gentoo-user] Re: GCC-4.3.2

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 897 bytes --]

On Mon, 6 Apr 2009 14:28:40 -0600, Joseph wrote:

> If you have look for it more carefully you would have found similar
> recommendation on the official gentoo pages:
> http://www.gentoo.org/doc/en/distcc.xml I think the above page qualify
> as an official doc. isn't it? 

Yes.

> "...-march=native
> 
> Starting with GCC 4.3.0, the compiler supports the -march=native switch
> which turns on CPU autodetection and optimizations that are worth being
> enabled on the processor the GCC is running at..."
> 
> So reading it, it seems to me the "-march=native" is recommended as
> well.

There's no recommendation there, just a description of what it does.

one of which changes the fact that gentoo-wiki.com is not endorsed by
Gentoo and should not be represented as being official.


-- 
Neil Bothwick

If at first you don't succeed, give up. No use being a damn fool.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Re: GCC-4.3.2

[Jarry]

Nikos Chantziaras wrote:

> Why should I use "native" if I  know that my CPU is "athlon64" :P

Are you sure your cpu is not "athlon64-sse3"?
Look for pni in cat /proc/cpuinfo   :-)

Jarry

-- 
_______________________________________________________________
This mailbox accepts e-mails only from selected mailing-lists!
Everything else is considered to be spam and therefore deleted.

Re: [gentoo-user] Re: GCC-4.3.2

[Joseph]

On 04/06/09 18:53, Jarry wrote:
> Nikos Chantziaras wrote:
>
>> Why should I use "native" if I  know that my CPU is "athlon64" :P
>
> Are you sure your cpu is not "athlon64-sse3"?
> Look for pni in cat /proc/cpuinfo   :-)
>
> Jarry

Here is my cpuinfo; 

cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 47
model name      : AMD Athlon(tm) 64 Processor 3000+
stepping        : 0
cpu MHz         : 1802.243
cache size      : 512 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 
3dnow up rep_good nopl pni
bogomips        : 3604.48
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc

I'm runing 64-bit so in my case I think it will be Athlon not Sempron:

CHOST="x86_64-pc-linux-gnu"
CFLAGS="-march=k8 -O2 -pipe"

am I correct?

-- 
Joseph

[gentoo-user] Re: GCC-4.3.2

[Nikos Chantziaras]

Jarry wrote:
> Nikos Chantziaras wrote:
> 
>> Why should I use "native" if I  know that my CPU is "athlon64" :P
> 
> Are you sure your cpu is not "athlon64-sse3"?

That was just an example.  For my CPU I use "-march=core2" :)

Re: [gentoo-user] GCC-4.3.2

[Jorge Morais]

On Mon, 6 Apr 2009 00:02:21 -0600
Joseph <syscon780@gmail.com> wrote:

> On 04/03/09 10:34, alain.didierjean@free.fr wrote:
> >Has one of you guys already switched from gcc-4.1.2 to gcc-4.3.2 and
> >performed  "emerge system" ?
> >What gives ? Any problem ? Is it worth it right now ? Please tell...
> >
> >--
> > ~adj~
> 
> I just notice that there is a problem with CFLAGS -march=native on AMD64
> I recompile my system using this new flag and have a lot of problems. 
> an examples new xorg-server-1.5... would not compile, I switch back to my previous settings:
> -march=athlon64 and it compiled on first pass.
> So, now I'm recompiling the system with my old flag: -march=athlon64 :-/
> 
Have you filed a bug?
Since in a later message you say that you are not sure of whether you
should use -march=k8 or -march=k8-sse3, it seems that in your situation
-march=native is interesting.

-- 
Software is like sex: it is better when it is free. --Linus Torvalds

Re: [gentoo-user] GCC-4.3.2

[Joseph]

On 04/06/09 22:00, Jorge Morais wrote:
>On Mon, 6 Apr 2009 00:02:21 -0600
>> 
>> I just notice that there is a problem with CFLAGS -march=native on AMD64
>> I recompile my system using this new flag and have a lot of problems. 
>> an examples new xorg-server-1.5... would not compile, I switch back to my previous settings:
>> -march=athlon64 and it compiled on first pass.
>> So, now I'm recompiling the system with my old flag: -march=athlon64 :-/
>> 
>Have you filed a bug?
>Since in a later message you say that you are not sure of whether you
>should use -march=k8 or -march=k8-sse3, it seems that in your situation
>-march=native is interesting.

I'm not sure of anything, I'm getting deeper an deeper problems.
I've almost recompile the system with -march=k8 and the last package gcc-4.3.2 wouldn't compile, the session got stuck so I kill the session and couldn't 
start the restart the "emerge gcc" as the emerge from previous session lock the portage or gcc.
I boot from CD, chroot and tried to recompile the gcc but it has been compiling for over two hours and it can not finish, I think it is in some kind of 
loop mode?

-- 
Joseph

Re: [gentoo-user] GCC-4.3.2

[Mike Kazantsev]

[-- Attachment #1: Type: text/plain, Size: 657 bytes --]

On Mon, 6 Apr 2009 20:19:09 -0600
Joseph <syscon780@gmail.com> wrote:

> I boot from CD, chroot and tried to recompile the gcc but it has been compiling for over two hours and it can not finish, I think it is in some kind of 
> loop mode?

GCC takes a long time to compile and re-compile itself with itself, but
since you experience it to be longer than usual, you probably have a
different settings, like "-jX" in MAKEOPTS or trying to compile it in
framebuffer, which takes longer, in my experience, I assume it might be
related to very slow echo operations, or might be just a glitch of my
perception ;)

-- 
Mike Kazantsev // fraggod.net

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] GCC-4.3.2

[Arttu V.]

On 4/7/09, Joseph <syscon780@gmail.com> wrote:
> I boot from CD, chroot and tried to recompile the gcc but it has been
> compiling for over two hours and it can not finish, I think it is in some
> kind of  loop mode?

That "loop" is the multilib profile you're probably using, which will
compile you both 64-bit and 32-bit version of gcc?

You also stated in an earlier email that your system has an Athlon
3000+ processor. As a comparison, here with an Athlon 3500+ and 3GB of
RAM the gcc:4.3 compilation takes around two hours if the ccache is
empty and an hour and 20 minutes if it hits mostly the cache, so the
duration doesn't appear surprising for a first compilation (varying by
the USE flags, naturally).

-- 
Arttu V.

Re: [gentoo-user] GCC-4.3.2

[Joseph]

On 04/07/09 19:48, Arttu V. wrote:
>On 4/7/09, Joseph <syscon780@gmail.com> wrote:
>> I boot from CD, chroot and tried to recompile the gcc but it has been
>> compiling for over two hours and it can not finish, I think it is in some
>> kind of  loop mode?
>
>That "loop" is the multilib profile you're probably using, which will
>compile you both 64-bit and 32-bit version of gcc?
>
>You also stated in an earlier email that your system has an Athlon
>3000+ processor. As a comparison, here with an Athlon 3500+ and 3GB of
>RAM the gcc:4.3 compilation takes around two hours if the ccache is
>empty and an hour and 20 minutes if it hits mostly the cache, so the
>duration doesn't appear surprising for a first compilation (varying by
>the USE flags, naturally).
>
>-- 
>Arttu V.

I'm still fighting with my two machines, recompiling system, and world with new GCC even though, it was not necessary; I wish developers mentioned it 
after emerging new gcc "no need to recompile the system" 
I just read few posts on the forum and did what others suggested: recompiled the world but is is not going smooth; 
maybe it is time for me to try something easier "ubuntu etc" as I'm not very happy with Gentoo, the upgrade and installation are harder to manage. 

-- 
Joseph