Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[Bakul Shah]

> Oddly enough it only happens when linking hget...

Try

9 mk -k all |& grep -B 1 ^ld:

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[David Leimbach]

[-- Attachment #1: Type: text/plain, Size: 342 bytes --]

On 10/30/07, Bakul Shah <bakul+plan9@bitblocks.com> wrote:
>
> > Oddly enough it only happens when linking hget...
>
> Try
>
> 9 mk -k all |& grep -B 1 ^ld:
>


Yeah there's more than one failing here... Still haven't figured out what's
missing yet.

I'll check with some other apple developers who've had more time on Leopard.

[-- Attachment #2: Type: text/html, Size: 741 bytes --]

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[andrey mirtchovski]

I haven't gotten to the point where David is, but i'm trying (leopard
isn't "approved for use" in our organization yet, they expect it to be
in a few months, but i can still use it at home :).

something to add to the knowledge base (although pretty obvious, it
stunned me this morning): if you are upgrading a Tiger machine that
had xcode installed on it you must upgrade the xcode tools too, or you
won't be able to link any binary from p9p.

xcode is available on developer.apple.com or on the upgrade dvd that
has leopard.

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[David Leimbach]

[-- Attachment #1: Type: text/plain, Size: 873 bytes --]

On 10/31/07, andrey mirtchovski <mirtchovski@gmail.com> wrote:
>
> I haven't gotten to the point where David is, but i'm trying (leopard
> isn't "approved for use" in our organization yet, they expect it to be
> in a few months, but i can still use it at home :).
>
> something to add to the knowledge base (although pretty obvious, it
> stunned me this morning): if you are upgrading a Tiger machine that
> had xcode installed on it you must upgrade the xcode tools too, or you
> won't be able to link any binary from p9p.
>
> xcode is available on developer.apple.com or on the upgrade dvd that
> has leopard.



I'm not sure I'd trust Apple's upgrading at all :-).

I just backup to another disk, then clean/install.


Then I may have some programs to fetch, and my preferences might be a little
out of whack, but I find less "surprises".

Dave

[-- Attachment #2: Type: text/html, Size: 1479 bytes --]

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[andrey mirtchovski]

the problem comes from libmp: the machine-dependent files written in
assembly (/src/libmp/386/*-Darwin.s) are causing it (and they are all
broken, not just the one reported). it's beyond me to figure out why
at the moment. i tried a couple of things at random, but that's not a
way to fix a problem.

all the functions defined therein have a replacement written in C in
the port/ directory of libmp. a workaround for now will be to fix the
mkfile in port/ to not look for the machine-dependent files (comment
out the test in the reduce script and let it just echo all file
names).

next bug is all the way to venti:

9l -o o.buildindex buildindex.o libvs.a /Users/andrey/plan9/lib/libventi.a
ld: in libvs.a, archive has no table of contents
collect2: ld returned 1 exit status

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[Russ Cox]

> next bug is all the way to venti:
>
> 9l -o o.buildindex buildindex.o libvs.a /Users/andrey/plan9/lib/libventi.a
> ld: in libvs.a, archive has no table of contents
> collect2: ld returned 1 exit status

looks like you might have to update
$PLAN9/bin/9ar if uname is printing
something other than "Darwin" these days.

russ

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[David Leimbach]

On Oct 31, 2007 12:05 PM, andrey mirtchovski <mirtchovski@gmail.com> wrote:
> the problem comes from libmp: the machine-dependent files written in
> assembly (/src/libmp/386/*-Darwin.s) are causing it (and they are all
> broken, not just the one reported). it's beyond me to figure out why
> at the moment. i tried a couple of things at random, but that's not a
> way to fix a problem.
>
> all the functions defined therein have a replacement written in C in
> the port/ directory of libmp. a workaround for now will be to fix the
> mkfile in port/ to not look for the machine-dependent files (comment
> out the test in the reduce script and let it just echo all file
> names).
>
> next bug is all the way to venti:
>
> 9l -o o.buildindex buildindex.o libvs.a /Users/andrey/plan9/lib/libventi.a
> ld: in libvs.a, archive has no table of contents
> collect2: ld returned 1 exit status
>
ranlib to the rescue... Got it run it explicitly on Darwin.  I think
gcc lets you get away without it on linux.

Dave

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[David Leimbach]

On Oct 31, 2007 1:18 PM, Russ Cox <rsc@swtch.com> wrote:
> > next bug is all the way to venti:
> >
> > 9l -o o.buildindex buildindex.o libvs.a /Users/andrey/plan9/lib/libventi.a
> > ld: in libvs.a, archive has no table of contents
> > collect2: ld returned 1 exit status
>
> looks like you might have to update
> $PLAN9/bin/9ar if uname is printing
> something other than "Darwin" these days.
>
> russ
>

uname still says "Darwin"

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[Michaelian Ennis]

On Oct 31, 2007, at 12:42 PM, David Leimbach wrote:
>
> I'm not sure I'd trust Apple's upgrading at all :-).
>
> I just backup to another disk, then clean/install.
>

I've done the upgrade six hosts with varied configurations with no  
real side effects detected.  Two hosts have been in use constantly  
with a few rosetta apps (adobe) being used regularly since the night  
of the release. I still backed my stuff up first just in case.

Apple Mail needed to be configured not to store deleted messages and  
drafts on the plan 9 IMAP service though as the folders it used had  
spaces in the names.

Ian

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[Michaelian Ennis]

d
On Oct 31, 2007, at 4:33 PM, David Leimbach wrote:

> On Oct 31, 2007 1:18 PM, Russ Cox <rsc@swtch.com> wrote:
>>> next bug is all the way to venti:
>>>
>>> 9l -o o.buildindex buildindex.o libvs.a /Users/andrey/plan9/lib/ 
>>> libventi.a
>>> ld: in libvs.a, archive has no table of contents
>>> collect2: ld returned 1 exit status
>>
>> looks like you might have to update
>> $PLAN9/bin/9ar if uname is printing
>> something other than "Darwin" these days.
>>
>> russ
>>
>
> uname still says "Darwin"

specifically (before devtools were installed):
: india ; uname -a
Darwin india.coraid.com 9.0.0 Darwin Kernel Version 9.0.0: Tue Oct  9  
21:35:55 PDT 2007; root:xnu-1228~1/RELEASE_I386 i386
: india ; gcc -v
Using built-in specs.
Target: i686-apple-darwin8
Configured with: /private/var/tmp/gcc/gcc-5367.obj~1/src/configure -- 
disable-checking -enable-werror --prefix=/usr --mandir=/share/man -- 
enable-languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg] 
[^.-]*$/s/$/-4.0/ --with-gxx-include-dir=/include/c++/4.0.0 --with- 
slibdir=/usr/lib --build=powerpc-apple-darwin8 --with-arch=nocona -- 
with-tune=generic --program-prefix= --host=i686-apple-darwin8 -- 
target=i686-apple-darwin8
Thread model: posix
gcc version 4.0.1 (Apple Computer, Inc. build 5367)
: india ;

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[Michaelian Ennis]

After dev tools install it still breaks in the same place but the  
compiler has different default settings now.  Here they are.

: india ; gcc -v
Using built-in specs.
Target: i686-apple-darwin9
Configured with: /var/tmp/gcc/gcc-5465~16/src/configure --disable- 
checking -enable-werror --prefix=/usr --mandir=/share/man --enable- 
languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/ 
$/-4.0/ --with-gxx-include-dir=/include/c++/4.0.0 --with-slibdir=/usr/ 
lib --build=i686-apple-darwin9 --with-arch=apple --with-tune=generic -- 
host=i686-apple-darwin9 --target=i686-apple-darwin9
Thread model: posix
gcc version 4.0.1 (Apple Inc. build 5465)

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[underspecified]

[-- Attachment #1: Type: text/plain, Size: 1307 bytes --]

Greetings,
Could you tell me how exactly you called ranlib to fix this problem?
I'm getting the same errors when compiling Acme SAC under leopard.

--underspecified

On 11/1/07, David Leimbach <leimy2k@gmail.com> wrote:
>
> On Oct 31, 2007 12:05 PM, andrey mirtchovski <mirtchovski@gmail.com>
> wrote:
> > the problem comes from libmp: the machine-dependent files written in
> > assembly (/src/libmp/386/*-Darwin.s) are causing it (and they are all
> > broken, not just the one reported). it's beyond me to figure out why
> > at the moment. i tried a couple of things at random, but that's not a
> > way to fix a problem.
> >
> > all the functions defined therein have a replacement written in C in
> > the port/ directory of libmp. a workaround for now will be to fix the
> > mkfile in port/ to not look for the machine-dependent files (comment
> > out the test in the reduce script and let it just echo all file
> > names).
> >
> > next bug is all the way to venti:
> >
> > 9l -o o.buildindex buildindex.o libvs.a/Users/andrey/plan9/lib/libventi.a
> > ld: in libvs.a, archive has no table of contents
> > collect2: ld returned 1 exit status
> >
> ranlib to the rescue... Got it run it explicitly on Darwin.  I think
> gcc lets you get away without it on linux.
>
> Dave
>

[-- Attachment #2: Type: text/html, Size: 1819 bytes --]

Re: [9fans] Re: Plan 9 Port error in mpvecadd-Darwin.s on Leopard

[underspecified]

[-- Attachment #1: Type: text/plain, Size: 1629 bytes --]

Sorry to reply to myself, but I got it working by adding a call to 'ranlib
$lib' after every call to ar.
Hope this helps anyone with similar problems.

--underspecified

On 11/3/07, underspecified <underspecified@gmail.com> wrote:
>
> Greetings,
>
> Could you tell me how exactly you called ranlib to fix this problem?
> I'm getting the same errors when compiling Acme SAC under leopard.
>
>
> --underspecified
>
> On 11/1/07, David Leimbach <leimy2k@gmail.com > wrote:
> >
> > On Oct 31, 2007 12:05 PM, andrey mirtchovski <mirtchovski@gmail.com >
> > wrote:
> > > the problem comes from libmp: the machine-dependent files written in
> > > assembly (/src/libmp/386/*-Darwin.s) are causing it (and they are all
> > > broken, not just the one reported). it's beyond me to figure out why
> > > at the moment. i tried a couple of things at random, but that's not a
> > > way to fix a problem.
> > >
> > > all the functions defined therein have a replacement written in C in
> > > the port/ directory of libmp. a workaround for now will be to fix the
> > > mkfile in port/ to not look for the machine-dependent files (comment
> > > out the test in the reduce script and let it just echo all file
> > > names).
> > >
> > > next bug is all the way to venti:
> > >
> > > 9l -o o.buildindex buildindex.o libvs.a/Users/andrey/plan9/lib/libventi.a
> > > ld: in libvs.a, archive has no table of contents
> > > collect2: ld returned 1 exit status
> > >
> > ranlib to the rescue... Got it run it explicitly on Darwin.  I think
> > gcc lets you get away without it on linux.
> >
> > Dave
> >
>
>
>

[-- Attachment #2: Type: text/html, Size: 2598 bytes --]