Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
GCC 3.1.1 Released
View unanswered posts
View posts from last 24 hours

 
Reply to topic    Gentoo Forums Forum Index Gentoo Chat
View previous topic :: View next topic  
Author Message
hubick
n00b
n00b


Joined: 23 Jul 2002
Posts: 8

PostPosted: Sat Jul 27, 2002 2:42 am    Post subject: GCC 3.1.1 Released Reply with quote

GCC 3.1.1 Released

3.2 will be out "within days or a week".
Back to top
View user's profile Send private message
Scandium
Retired Dev
Retired Dev


Joined: 22 Apr 2002
Posts: 340
Location: Germany

PostPosted: Sat Jul 27, 2002 5:35 am    Post subject: Reply with quote

Waiting for 3.2.1 :)
Back to top
View user's profile Send private message
hubick
n00b
n00b


Joined: 23 Jul 2002
Posts: 8

PostPosted: Sat Jul 27, 2002 6:24 am    Post subject: Reply with quote

Well, according to this, you should only have to wait until Sept 15.
Back to top
View user's profile Send private message
Scandium
Retired Dev
Retired Dev


Joined: 22 Apr 2002
Posts: 340
Location: Germany

PostPosted: Sat Jul 27, 2002 7:54 am    Post subject: Reply with quote

sure, but well...It'll take until 3.3.x before everybody drops 2.95.x...

But I am looking forward to 3.2.x because I believe it is generating useful code ;-)
Back to top
View user's profile Send private message
xorbe
n00b
n00b


Joined: 19 Jul 2002
Posts: 46
Location: near San Jose, CA

PostPosted: Sat Jul 27, 2002 8:04 am    Post subject: Reply with quote

Scandium wrote:
sure, but well...It'll take until 3.3.x before everybody drops 2.95.x...

No, 3.2.x is the stabilized 3.0/3.1/3.2 series, I believe.

3.3+ contains radical additions and changes. Besides,
doesn't the verson number 3.2.1 have a nice ring? :wink:
Back to top
View user's profile Send private message
Jesse
Tux's lil' helper
Tux's lil' helper


Joined: 24 Apr 2002
Posts: 148

PostPosted: Sat Jul 27, 2002 2:39 pm    Post subject: Reply with quote

eh actually 3.3 which was 3.2 isnt that radical in terms of 'stupid things we had to do over to get things to work cuz they were really really broken in the first place' kind of things ... But now for 3.4, they are rewritting the parser _again_ and mucking with the intermediate rtl I believe (Mark Mitchell and CodeSourcery)... that of course will mean more regressions than 3.0 had with 2.95 and then some ....

Of course, if it works out, the compilation speed should be vastly better, the optimization people will find it easier to use the new rtl for advanced forms of block optimizations and prefectching, the binary sizes may fall again :) AND that _lil_ problem where gcc eats all 3 of your swap drives compiling inline functions the proper way (a sane and intelligent way) will hopefully be gone.

because of this im just hoping that we dont have to wait for 3.5 before it's finally back on track with the reputation gcc has had in the past. But other than that rewrite and the other necessary stuff that has to get added yes 3.3 and 3.4 should both be fairly good.

However, if you are watching the nightly/weekly/sometimes monthly runs of Spec95 and Spec2000 and the Bench++ benchmarks being run on gcc.gnu.org, I cringe every time where I see 2.95.3 _still_ beating the crap outa 3.x and the ICC spec runs put gcc to shame on the runtime marks.
Back to top
View user's profile Send private message
Scandium
Retired Dev
Retired Dev


Joined: 22 Apr 2002
Posts: 340
Location: Germany

PostPosted: Sat Jul 27, 2002 3:53 pm    Post subject: Reply with quote

xorbe wrote:
Scandium wrote:
sure, but well...It'll take until 3.3.x before everybody drops 2.95.x...

No, 3.2.x is the stabilized 3.0/3.1/3.2 series, I believe.

3.3+ contains radical additions and changes. Besides,
doesn't the verson number 3.2.1 have a nice ring? :wink:


But anyway you already said it...It's 3.3 which has the big impact and not 3.2 (over 3.1).
Back to top
View user's profile Send private message
xorbe
n00b
n00b


Joined: 19 Jul 2002
Posts: 46
Location: near San Jose, CA

PostPosted: Sat Jul 27, 2002 5:41 pm    Post subject: Reply with quote

Jesse wrote:
eh actually 3.3 which was 3.2 isnt that radical in terms of 'stupid things we had to do over to get things to work cuz they were really really broken in the first place' kind of things ...

Oh yeah, like the problem where they can't figure out whether it's a declaration or a statement due to some parenthesis cast and a function call, before it's too late.

function(casttype(data));

parses as a declaration, but they realize too late that it's actually a function call. :? I think my example it jacked, maybe someone who knows the nitty-gritty can explain the scenario properly...
Back to top
View user's profile Send private message
CYwolf
n00b
n00b


Joined: 01 Jun 2002
Posts: 47
Location: Toronto, ON, CA

PostPosted: Sat Jul 27, 2002 6:37 pm    Post subject: Erm, that's a cast? Reply with quote

I could see that as a call to a class constructor maybe, but most people do casts via
Code:
function((casttype)data);

:?
Back to top
View user's profile Send private message
Jesse
Tux's lil' helper
Tux's lil' helper


Joined: 24 Apr 2002
Posts: 148

PostPosted: Sat Jul 27, 2002 7:12 pm    Post subject: Reply with quote

Yeah, thats a _normal_ ''stupid thing we had to do over to get things to work cuz they were really really broken in the first place" thing. There's usually quite a bit of those. Not nearly as 'radical' as a parser rewrite though.

Examples of major changes include:

A new C++ parser
A new memory allocation scheme for internal use in the compiler
A new optimization pass


Hmmm, I believe there was a parsing problem similar to this on the gcc mailing list a few days back .... http://gcc.gnu.org/ml/gcc/2002-07/msg01033.html

Is your problem known?
Back to top
View user's profile Send private message
xorbe
n00b
n00b


Joined: 19 Jul 2002
Posts: 46
Location: near San Jose, CA

PostPosted: Sat Jul 27, 2002 9:38 pm    Post subject: Reply with quote

Jesse wrote:
Hmmm, I believe there was a parsing problem similar to this on the gcc mailing list a few days back .... http://gcc.gnu.org/ml/gcc/2002-07/msg01033.html

Is your problem known?

Yeah, that's the example from the mailing list that I was thinking about. Not my problem! :o

It's quite interesting to read through the gcc bugs... some are silly, some are nastily complex...
Back to top
View user's profile Send private message
rac
Bodhisattva
Bodhisattva


Joined: 30 May 2002
Posts: 6553
Location: Japanifornia

PostPosted: Mon Aug 19, 2002 2:13 am    Post subject: Reply with quote

GCC3 to Gentoo Chat.
_________________
For every higher wall, there is a taller ladder
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Gentoo Chat All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum