alternative compilers

alternative compilers

[ron minnich]

fyi, you can get nix, but you can't get a real 64-bit compiler for
AMD. Why am I saying this? Isn't amd64 64-bit?

Yes, but ... 6c/6l don't produce code that can use 64 bit pointers.
The pointers are always truncated to 32 bits, so you have to use
pointers that sign extend, i.e.
kernel mode pointers have to be ffffffff'80000000 or greater, and user
pointers have to be 7fffffff or less. Yes, I'm well aware that stack
is at 7fff ffffffff but that's actually a special case.

So, question, anyone interested in making other compilers work on
9front? If so, which ones?

you may find this hard to  believe. If you do, take this code:
#include <u.h>
#include <libc.h>
char data = 5;
b(void){
	 char *c = &data ;
	return *c * 5;
}
main()
{
	int i;
	print("&i %#p\n", &i);
	print("Good morning\n");
	print("%d\n", b());
}
/* 6c b.c; 6l -R0x20000000 b.6; nm -n 6.out */

and run this command:
6l -a -R0x20000000 b.6

and looked at the code for b(). Then come back and talk :-)

ron
p.s. Of course, another option is to modify 6l to do better. That
would be cool too. But at the same time, I'm not so sure how long this
should go on. Maybe we should use this one: http://bellard.org/tcc/

Re: alternative compilers

[Julius Schmidt]

> Yes, but ... 6c/6l don't produce code that can use 64 bit pointers.
> The pointers are always truncated to 32 bits, so you have to use
> pointers that sign extend, i.e.
> kernel mode pointers have to be ffffffff'80000000 or greater, and user
> pointers have to be 7fffffff or less. Yes, I'm well aware that stack
> is at 7fff ffffffff but that's actually a special case.

So 6c/6l is an elaborate april fools' joke?
What other explanation is there?
I mean what the fuck, a 64 bit compiler which doesn't do 64 bit
pointers?

> So, question, anyone interested in making other compilers work on
> 9front? If so, which ones?

I'm not entirely happy with adding 64 bit support.
There are many PC hardware issues with 64 bits, it's quite a bit of work
and that just for more accessible RAM?
I can't even fill 2 GB. ;P

I'd prefer focussing on getting hardware working and all that other stuff
we want to do.

I don't know of any worthy compiler:
- gcc: FOR THE LOVE OF JESUS NO
- pcc: 80s want their compiler back (does it even have AMD64 support?)
- clang: C++; also just gcc in green
- tcc: might be even worth a shot. no clue if it is sophisticated enough
   to compile plan 9 code

The only option, if any, is fixing kencc, I'd say.

Re: alternative compilers

[ron minnich]

On Mon, Jun 20, 2011 at 2:49 PM, Julius Schmidt <aiju@phicode.de> wrote:

> The only option, if any, is fixing kencc, I'd say.
>

ok, just an fyi, the code as usual is very good (ken + charles ==
good), but did wanted you all to know lest there be surprises.

Plus, maybe one of you would enjoy the challenge.

ron

Re: alternative compilers

[Julius Schmidt]

> ok, just an fyi, the code as usual is very good (ken + charles ==
> good), but did wanted you all to know lest there be surprises.
>
> Plus, maybe one of you would enjoy the challenge.
>
> ron

Already busy with fixing 5c :)
http://code.google.com/p/plan9front/issues/detail?id=58
(I aspire to do that one once I have a pandaboard)

Re: alternative compilers

[suharik]

> - clang: C++; also just gcc in green
Is it possible to simplify llvm/clang compiler?

Re: alternative compilers

[Noah Evans]

The minix guys just did a bunch of work incorporating clang et al into
the minix codebase, they might have interesting opinions on the
matter, if you guys are interested.

Noah



On Tue, Jun 21, 2011 at 12:01 PM, suharik <gleb.ax.sh@gmail.com> wrote:
>> - clang: C++; also just gcc in green
> Is it possible to simplify llvm/clang compiler?
>

Re: alternative compilers

[Uriel]

On Tue, Jun 21, 2011 at 12:31 PM, Noah Evans <noah.evans@gmail.com> wrote:
> The minix guys just did a bunch of work incorporating clang et al into
> the minix codebase, they might have interesting opinions on the
> matter, if you guys are interested.

Interested in a C++ project that uses GCC as its backend? Are you drunk?

uriel

> Noah
>
>
>
> On Tue, Jun 21, 2011 at 12:01 PM, suharik <gleb.ax.sh@gmail.com> wrote:
>>> - clang: C++; also just gcc in green
>> Is it possible to simplify llvm/clang compiler?
>

Re: alternative compilers

[Noah Evans]

http://www.amazon.com/How-Win-Friends-Influence-People/dp/0671723650

Noah



On Tue, Jun 21, 2011 at 1:13 PM, Uriel <uriel@berlinblue.org> wrote:
> On Tue, Jun 21, 2011 at 12:31 PM, Noah Evans <noah.evans@gmail.com> wrote:
>> The minix guys just did a bunch of work incorporating clang et al into
>> the minix codebase, they might have interesting opinions on the
>> matter, if you guys are interested.
>
> Interested in a C++ project that uses GCC as its backend? Are you drunk?
>
> uriel
>
>> Noah
>>
>>
>>
>> On Tue, Jun 21, 2011 at 12:01 PM, suharik <gleb.ax.sh@gmail.com> wrote:
>>>> - clang: C++; also just gcc in green
>>> Is it possible to simplify llvm/clang compiler?
>>
>

Re: alternative compilers

[suharik]

Anyway, we forget about open watcom.

Re: alternative compilers

[Julius Schmidt]

> http://www.amazon.com/How-Win-Friends-Influence-People/dp/0671723650
>
> Noah

How to be a cunt on the Internet and troll 9front developers by Noah
Evans.

Amen to uriel.

Re: alternative compilers

[Noah Evans]

Temper temper. I was serious BTW, the minix guys are quite nice and
have given the compiler issue a lot of thought.

Noah



On Tue, Jun 21, 2011 at 3:56 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> http://www.amazon.com/How-Win-Friends-Influence-People/dp/0671723650
>>
>> Noah
>
> How to be a cunt on the Internet and troll 9front developers by Noah
> Evans.
>
> Amen to uriel.
>

Re: alternative compilers

[Julius Schmidt]

> Temper temper. I was serious BTW, the minix guys are quite nice and
> have given the compiler issue a lot of thought.

The road to hell is paved with good intentions.
Especially true of software.

Re: alternative compilers

[Noah Evans]

You pick your battles. When you're part of the infinite monkey brigade
you can afford to be inflexible. When you have a paper or grant
deadline looming you'll find yourself a bit more pragmatic.

Noah



On Tue, Jun 21, 2011 at 4:47 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> Temper temper. I was serious BTW, the minix guys are quite nice and
>> have given the compiler issue a lot of thought.
>
> The road to hell is paved with good intentions.
> Especially true of software.
>

Re: alternative compilers

[Iruatã Souza]

On Tue, Jun 21, 2011 at 11:50 AM, Noah Evans <noah.evans@gmail.com> wrote:
> You pick your battles. When you're part of the infinite monkey brigade
> you can afford to be inflexible. When you have a paper or grant
> deadline looming you'll find yourself a bit more pragmatic.
>
> Noah
>

Oddly enough, we have no deadlines to meet and still we are solving
very pragmatic (and often ignored) problems.

>
> On Tue, Jun 21, 2011 at 4:47 PM, Julius Schmidt <aiju@phicode.de> wrote:
>>> Temper temper. I was serious BTW, the minix guys are quite nice and
>>> have given the compiler issue a lot of thought.
>>
>> The road to hell is paved with good intentions.
>> Especially true of software.
>>
>

Re: alternative compilers

[Noah Evans]

Good for you. Really. However, these solutions are worthless unless
you a. build a community inclusive enough for people to use them or b.
get them merged back into the main tree.

Noah



On Wed, Jun 22, 2011 at 12:21 AM, Iruatã Souza <iru.muzgo@gmail.com> wrote:
> On Tue, Jun 21, 2011 at 11:50 AM, Noah Evans <noah.evans@gmail.com> wrote:
>> You pick your battles. When you're part of the infinite monkey brigade
>> you can afford to be inflexible. When you have a paper or grant
>> deadline looming you'll find yourself a bit more pragmatic.
>>
>> Noah
>>
>
> Oddly enough, we have no deadlines to meet and still we are solving
> very pragmatic (and often ignored) problems.
>
>>
>> On Tue, Jun 21, 2011 at 4:47 PM, Julius Schmidt <aiju@phicode.de> wrote:
>>>> Temper temper. I was serious BTW, the minix guys are quite nice and
>>>> have given the compiler issue a lot of thought.
>>>
>>> The road to hell is paved with good intentions.
>>> Especially true of software.
>>>
>>
>

Re: alternative compilers

[Julius Schmidt]

> Good for you. Really. However, these solutions are worthless unless
> you a. build a community inclusive enough for people to use them or b.
> get them merged back into the main tree.

c. We use them. That's worthwhile enough.

Historically, languages designed for other people to use have been bad:
Cobol, PL/I, Pascal, Ada, C++. The good languages have been those that
were designed for their own creators: C, Perl, Smalltalk, Lisp. -- Paul
Graham

I'm about to make lunch.
Should I build a community around it or merge it back into the main tree
to prevent it from being worthless?

Re: alternative compilers

[Noah Evans]

My goodness it can be (almost) civil.

You bring up an interesting point. Plan9 was designed for Ken's music
collection(Unix was likewise created for space travel). However the
reason that Unix succeeded was that its creators were willing to allow
the system to be used to meet *other* people's needs. It's not enough
to scratch your own itch, you have to be cooperative enough to allow
others to contribute as well. I find this attitude especially ironic
given that, to the best of my knowledge, 9front was established based
on the belief that certain maintainers were not being responsive
enough to the needs of the community.

Noah



On Wed, Jun 22, 2011 at 12:04 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> Good for you. Really. However, these solutions are worthless unless
>> you a. build a community inclusive enough for people to use them or b.
>> get them merged back into the main tree.
>
> c. We use them. That's worthwhile enough.
>
> Historically, languages designed for other people to use have been bad:
> Cobol, PL/I, Pascal, Ada, C++. The good languages have been those that
> were designed for their own creators: C, Perl, Smalltalk, Lisp. -- Paul
> Graham
>
> I'm about to make lunch.
> Should I build a community around it or merge it back into the main tree
> to prevent it from being worthless?
>

Re: alternative compilers

[Julius Schmidt]

> You bring up an interesting point. Plan9 was designed for Ken's music
> collection(Unix was likewise created for space travel). However the
> reason that Unix succeeded was that its creators were willing to allow
> the system to be used to meet *other* people's needs. It's not enough
> to scratch your own itch, you have to be cooperative enough to allow
> others to contribute as well. I find this attitude especially ironic
> given that, to the best of my knowledge, 9front was established based
> on the belief that certain maintainers were not being responsive
> enough to the needs of the community.

I'm not exactly sure what you're contemplating but we're not going to
use
1. BSD code
2. Minix code
3. anything in C++
(this list can be extended indefinitely)

I'm not alone with that opinion in the 9front team.
If you don't like it, go somewhere else.

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 1:06 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> You bring up an interesting point. Plan9 was designed for Ken's music
>> collection(Unix was likewise created for space travel). However the
>> reason that Unix succeeded was that its creators were willing to allow
>> the system to be used to meet *other* people's needs. It's not enough
>> to scratch your own itch, you have to be cooperative enough to allow
>> others to contribute as well. I find this attitude especially ironic
>> given that, to the best of my knowledge, 9front was established based
>> on the belief that certain maintainers were not being responsive
>> enough to the needs of the community.
>
> I'm not exactly sure what you're contemplating but we're not going to
> use
> 1. BSD code

I have to partially disagree here, for example integrating some bits
of OpenBSD's libc and core utils into APE might make sense, as might
in some rare cases make sense to use some bits of their drivers.

But in general, yea, we got the best C compilers in the world, why the
hell would anyone want to replace them with a bunch of putrid
convoluted C++!?!

uriel

Re: alternative compilers

[Julius Schmidt]

> I have to partially disagree here, for example integrating some bits
> of OpenBSD's libc and core utils into APE might make sense, as might
> in some rare cases make sense to use some bits of their drivers.
The exception proves the rule.

Re: alternative compilers

[Noah Evans]

Yawn... one of these days you'll understand. Ta ta.

Noah



On Wed, Jun 22, 2011 at 1:06 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> You bring up an interesting point. Plan9 was designed for Ken's music
>> collection(Unix was likewise created for space travel). However the
>> reason that Unix succeeded was that its creators were willing to allow
>> the system to be used to meet *other* people's needs. It's not enough
>> to scratch your own itch, you have to be cooperative enough to allow
>> others to contribute as well. I find this attitude especially ironic
>> given that, to the best of my knowledge, 9front was established based
>> on the belief that certain maintainers were not being responsive
>> enough to the needs of the community.
>
> I'm not exactly sure what you're contemplating but we're not going to
> use
> 1. BSD code
> 2. Minix code
> 3. anything in C++
> (this list can be extended indefinitely)
>
> I'm not alone with that opinion in the 9front team.
> If you don't like it, go somewhere else.
>

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 1:15 PM, Noah Evans <noah.evans@gmail.com> wrote:
> Yawn... one of these days you'll understand. Ta ta.

You should learn to be a bit more condescending. I'm sure that will
help you gain even more "friends" and influence.

uriel

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 12:49 PM, Noah Evans <noah.evans@gmail.com> wrote:
> My goodness it can be (almost) civil.
>
> You bring up an interesting point. Plan9 was designed for Ken's music
> collection(Unix was likewise created for space travel). However the
> reason that Unix succeeded was that its creators were willing to allow
> the system to be used to meet *other* people's needs.
> It's not enough to scratch your own itch, you have to be cooperative enough to allow
> others to contribute as well.

I almost fell off my chair when I read this. Plan 9 development might
be many things, but it is not cooperative.

I have seen more cooperation in the last few weeks in #cat-v than I
have seen in 9fans for years (but then I stopped reading 9fans a while
ago, it is an even bigger waste of time than reading the random
obscene jokes in #cat-v, at least some of those are funny).

uriel

> I find this attitude especially ironic
> given that, to the best of my knowledge, 9front was established based
> on the belief that certain maintainers were not being responsive
> enough to the needs of the community.
>
> Noah
>
>
>
> On Wed, Jun 22, 2011 at 12:04 PM, Julius Schmidt <aiju@phicode.de> wrote:
>>> Good for you. Really. However, these solutions are worthless unless
>>> you a. build a community inclusive enough for people to use them or b.
>>> get them merged back into the main tree.
>>
>> c. We use them. That's worthwhile enough.
>>
>> Historically, languages designed for other people to use have been bad:
>> Cobol, PL/I, Pascal, Ada, C++. The good languages have been those that
>> were designed for their own creators: C, Perl, Smalltalk, Lisp. -- Paul
>> Graham
>>
>> I'm about to make lunch.
>> Should I build a community around it or merge it back into the main tree
>> to prevent it from being worthless?
>>
>

Re: alternative compilers

[Iruatã Souza]

On Wed, Jun 22, 2011 at 7:49 AM, Noah Evans <noah.evans@gmail.com> wrote:
> My goodness it can be (almost) civil.
>
> You bring up an interesting point. Plan9 was designed for Ken's music
> collection(Unix was likewise created for space travel). However the
> reason that Unix succeeded was that its creators were willing to allow
> the system to be used to meet *other* people's needs. It's not enough
> to scratch your own itch, you have to be cooperative enough to allow
> others to contribute as well. I find this attitude especially ironic
> given that, to the best of my knowledge, 9front was established based
> on the belief that certain maintainers were not being responsive
> enough to the needs of the community.
>

You might have a wrong belief, then. I don't think we ever talked
about community.
We are just a bunch of people who like Plan 9 and are changing it in
our own ways.
How big a problem is that?

iru

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 2:10 PM, Iruatã Souza <iru.muzgo@gmail.com> wrote:
> On Wed, Jun 22, 2011 at 7:49 AM, Noah Evans <noah.evans@gmail.com> wrote:
>> My goodness it can be (almost) civil.
>>
>> You bring up an interesting point. Plan9 was designed for Ken's music
>> collection(Unix was likewise created for space travel). However the
>> reason that Unix succeeded was that its creators were willing to allow
>> the system to be used to meet *other* people's needs. It's not enough
>> to scratch your own itch, you have to be cooperative enough to allow
>> others to contribute as well. I find this attitude especially ironic
>> given that, to the best of my knowledge, 9front was established based
>> on the belief that certain maintainers were not being responsive
>> enough to the needs of the community.
>>
>
> You might have a wrong belief, then. I don't think we ever talked
> about community.
>
> We are just a bunch of people who like Plan 9 and are changing it in
> our own ways.
> How big a problem is that?

It is a problem for some people when it exposes the dysfunctional Plan
9 "community" for the joke that it is.

uriel

Re: alternative compilers

[Julius Schmidt]

> It is a problem for some people when it exposes the dysfunctional Plan
> 9 "community" for the joke that it is.

You seem to implying that there is only one Plan 9 community.

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 2:41 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> It is a problem for some people when it exposes the dysfunctional Plan
>> 9 "community" for the joke that it is.
>
> You seem to implying that there is only one Plan 9 community.

The quotes where there for a reason, I didn't mean to imply that there
are *any* Plan 9 communities at all.

uriel

Re: alternative compilers

[Stanley Lieber]

There is no problem here. No one is required to use (or not use) 9front.

"You have the source." -Russ Cox

-sl

Re: alternative compilers

[John Floren]

On Wed, Jun 22, 2011 at 7:54 AM, Stanley Lieber
<stanley.lieber@gmail.com> wrote:
> There is no problem here. No one is required to use (or not use) 9front.
>
> "You have the source." -Russ Cox
>
> -sl
>

If someone was too disgusted with the 9front people, he could fork 9front too...


John

Re: alternative compilers

[Stanley Lieber]

> If someone was too disgusted with the 9front people, he could fork 9front too...

Of course.

Why not?

-sl

Re: alternative compilers

[Kurt H Maier]

> I find this attitude especially ironic
> given that, to the best of my knowledge, 9front was established based
> on the belief that certain maintainers were not being responsive
> enough to the needs of the community.

9front was established based on the belief that certain maintainers
were not being responsive enough to the needs of the creators of
9front.

Re: alternative compilers

[Stanley Lieber]

http://thegreen.stanleylieber.com/src/676962418.png

-sl

Re: alternative compilers

[cinap_lenrek]

no, wrong!

it was created to play DooM all along!

--
cinap

Re: alternative compilers

[Kurt H Maier]

On Sun, Jun 26, 2011 at 8:34 PM,  <cinap_lenrek@gmx.de> wrote:
> no, wrong!
>
> it was created to play DooM all along!

I already said that!


-- 
# Kurt H Maier

Re: alternative compilers

[Uriel]

On Wed, Jun 22, 2011 at 12:04 PM, Julius Schmidt <aiju@phicode.de> wrote:
>
> I'm about to make lunch.
> Should I build a community around it or merge it back into the main tree
> to prevent it from being worthless?

9front development is already more community driven than Plan 9 ever was.

We have a cat-v.org community, not a politically correct community,
but still a healthy community of dedicated and intelligent (not me)
people with a working sense of humour.

Of course people that are more interested in feeling offended by
stupid words than on getting working stuff out will stay away from
this community, that is fine, look at how much 'good' they have done
to the Plan 9 "community".

Somehow a bunch of offensive crazy trolls have managed to gather and
integrate more (useful) contributions in a few weeks than Bell Labs
and the PC-gestapo in 9fans have managed in years.

So please, tell us more about how to create and run a community, so we
can make sure to do exactly the opposite.

uriel

P.S.: My only contributions to 9front so far have been my regular
trolling and trying to keep people from doing too-insanely-stupid shit
like fucking around with the 9P protocol, all credit should obviously
go to all the people kicking ass and writing code and documentation,
who have proved me wrong (once again) when I said a fork of Plan 9
would never be successful.

Re: alternative compilers

[Uriel]

On Tue, Jun 21, 2011 at 4:47 PM, Julius Schmidt <aiju@phicode.de> wrote:
>> Temper temper. I was serious BTW, the minix guys are quite nice and
>> have given the compiler issue a lot of thought.
>
> The road to hell is paved with good intentions.
> Especially true of software.

Good intentions and features.

But the minix guys are really nice, they even managed to reduce their
kernel to one syscall, ioctl.

uriel

Re: alternative compilers

[Uriel]

On Tue, Jun 21, 2011 at 12:01 PM, suharik <gleb.ax.sh@gmail.com> wrote:
>> - clang: C++; also just gcc in green
> Is it possible to simplify llvm/clang compiler?

No.

uriel

Re: alternative compilers

[Jacob Todd]

[-- Attachment #1: Type: text/plain, Size: 159 bytes --]

No.
On Jun 21, 2011 6:01 AM, "suharik" <gleb.ax.sh@gmail.com> wrote:
>> - clang: C++; also just gcc in green
> Is it possible to simplify llvm/clang compiler?

[-- Attachment #2: Type: text/html, Size: 315 bytes --]

Re: alternative compilers

[Jens Staal]

> So, question, anyone interested in making other compilers work on
> 9front? If so, which ones?
>

I am currently playing with trying to get PCC (the Portable kind, not
the APE kind) to compile under APE. I am however stuck already at the
first step of getting past autoconf hell (mkmk does not seem to know
what to do and I am a noob)

(A cool thing I was thinking aobut for PCC on plan9 was actually to
bring up an alternative *BSD kernel (FreeBSD kernel can apparently be
compiled with PCC) with glendix-type patches natively on Plan9. This
"glendie" could thus be an alternative kernel for extended hardware
support)

Re: alternative compilers

[Iruatã Souza]

On Tue, Jun 21, 2011 at 1:08 AM, Jens Staal <staal1978@gmail.com> wrote:
>> So, question, anyone interested in making other compilers work on
>> 9front? If so, which ones?
>>
>
> I am currently playing with trying to get PCC (the Portable kind, not
> the APE kind) to compile under APE. I am however stuck already at the
> first step of getting past autoconf hell (mkmk does not seem to know
> what to do and I am a noob)
>
> (A cool thing I was thinking aobut for PCC on plan9 was actually to
> bring up an alternative *BSD kernel (FreeBSD kernel can apparently be
> compiled with PCC) with glendix-type patches natively on Plan9. This
> "glendie" could thus be an alternative kernel for extended hardware
> support)
>

thanks, but we are writing our own drivers.

Re: alternative compilers

[Jens Staal]

>> (A cool thing I was thinking aobut for PCC on plan9 was actually to
>> bring up an alternative *BSD kernel (FreeBSD kernel can apparently be
>> compiled with PCC) with glendix-type patches natively on Plan9. This
>> "glendie" could thus be an alternative kernel for extended hardware
>> support)
>>
>
> thanks, but we are writing our own drivers.
>

I do not see an inherent conflict between those two things. This was
also not the main reason to try to get PCC working on Plan9 (I tend to
speculate wildly sometimes - life gets more fun that way). I first of
all just thought it was the most likely one to be possible to port and
which might bring some benefit (TCC would be nice too, but that one
has portability issues even to BSD last thing I read).

Anyone know anything about open64/ecopath64?

Re: alternative compilers

[Uriel]

On Tue, Jun 21, 2011 at 6:08 AM, Jens Staal <staal1978@gmail.com> wrote:
> (A cool thing I was thinking aobut for PCC on plan9 was actually to
> bring up an alternative *BSD kernel (FreeBSD kernel can apparently be
> compiled with PCC) with glendix-type patches natively on Plan9. This
> "glendie" could thus be an alternative kernel for extended hardware
> support)

If we wanted LSD we would be in Berkeley already.

If you want lunix, get a real lunix.

uriel

P.S.: OpenBSD is actually nice and useful, but makes much more sense
to port the Plan 9 compilers to obsd than the other way around.

Re: alternative compilers

[Uriel]

On Mon, Jun 20, 2011 at 11:39 PM, ron minnich <rminnich@gmail.com> wrote:
> fyi, you can get nix, but you can't get a real 64-bit compiler for
> AMD. Why am I saying this? Isn't amd64 64-bit?

Just wondering, does this apply to the 6c in the Go distribution too?
In any case, 6g does have real 64bit support, and we should be
switching as much code as possible to Go anyway.

uriel


> Yes, but ... 6c/6l don't produce code that can use 64 bit pointers.
> The pointers are always truncated to 32 bits, so you have to use
> pointers that sign extend, i.e.
> kernel mode pointers have to be ffffffff'80000000 or greater, and user
> pointers have to be 7fffffff or less. Yes, I'm well aware that stack
> is at 7fff ffffffff but that's actually a special case.
>
> So, question, anyone interested in making other compilers work on
> 9front? If so, which ones?
>
> you may find this hard to  believe. If you do, take this code:
> #include <u.h>
> #include <libc.h>
> char data = 5;
> b(void){
>         char *c = &data ;
>        return *c * 5;
> }
> main()
> {
>        int i;
>        print("&i %#p\n", &i);
>        print("Good morning\n");
>        print("%d\n", b());
> }
> /* 6c b.c; 6l -R0x20000000 b.6; nm -n 6.out */
>
> and run this command:
> 6l -a -R0x20000000 b.6
>
> and looked at the code for b(). Then come back and talk :-)
>
> ron
> p.s. Of course, another option is to modify 6l to do better. That
> would be cool too. But at the same time, I'm not so sure how long this
> should go on. Maybe we should use this one: http://bellard.org/tcc/
>