Eugene Kirpichov | 26 Sep 01:52 2012
Picon

64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Hi,

I installed Haskell Platform 32-bit on a fresh 64-bit mac, because the
page http://www.haskell.org/platform/mac.html says: "Pick the 32-bit
vesion, unless you have a specific reason to use the 64-bit version.
The 32-bit one is slightly faster for most programs."

Then, during the installation of a package, the following happeed:

Loading package cairo-0.12.3.1 ... <command line>: can't load .so/.DLL
for: /opt/local/lib/libz.dylib (dlopen(/opt/local/lib/libz.dylib, 9):
no suitable image found.  Did find:
        /opt/local/lib/libz.dylib: mach-o, but wrong architecture)

That libz.dylib is a 64-bit library and it can't be used by 32-bit
Haskell platform.

QUESTION: It seems to me that most people, at least most who care
about "slightly faster" programs, are likely to run into something
like this - using native 64-bit libraries. Compatibility exists only
in the opposite direction. Wouldn't it be appropriate to remove this
notice and ask people to use the 64-bit version unless they have a
specific reason not to?

--

-- 
Eugene Kirpichov
http://www.linkedin.com/in/eugenekirpichov
We're hiring! http://tinyurl.com/mirantis-openstack-engineer
Carter Schonwald | 26 Sep 07:44 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Hi Eugene,

To the best of my knowledge there is absolutely no reason to use the 32bit haskell on OS X (aside from memory usage optimization cases which likely do not matter to the *typical* user), and the community should probably update the recommendation to reflect this.

I can certainly attest to having to baby sit new haskellers when installing and repeatedly say  "yes, do the 64 bit version, please, no, ignore the recommendation for 32bit, no one knows why its there, but its wrong"

what can we (the community ) do to address the fact that the haskell platform installer suggestions for os x are sadly completely backwards? (or am I completely wrong in my personal stance on this matter)

cheers
-Carter 

On Tue, Sep 25, 2012 at 7:52 PM, Eugene Kirpichov <ekirpichov <at> gmail.com> wrote:
Hi,

I installed Haskell Platform 32-bit on a fresh 64-bit mac, because the
page http://www.haskell.org/platform/mac.html says: "Pick the 32-bit
vesion, unless you have a specific reason to use the 64-bit version.
The 32-bit one is slightly faster for most programs."

Then, during the installation of a package, the following happeed:

Loading package cairo-0.12.3.1 ... <command line>: can't load .so/.DLL
for: /opt/local/lib/libz.dylib (dlopen(/opt/local/lib/libz.dylib, 9):
no suitable image found.  Did find:
        /opt/local/lib/libz.dylib: mach-o, but wrong architecture)

That libz.dylib is a 64-bit library and it can't be used by 32-bit
Haskell platform.

QUESTION: It seems to me that most people, at least most who care
about "slightly faster" programs, are likely to run into something
like this - using native 64-bit libraries. Compatibility exists only
in the opposite direction. Wouldn't it be appropriate to remove this
notice and ask people to use the 64-bit version unless they have a
specific reason not to?


--
Eugene Kirpichov
http://www.linkedin.com/in/eugenekirpichov
We're hiring! http://tinyurl.com/mirantis-openstack-engineer

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Johan Tibell | 26 Sep 10:58 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Hi,

On Wed, Sep 26, 2012 at 7:44 AM, Carter Schonwald
<carter.schonwald <at> gmail.com> wrote:
> To the best of my knowledge there is absolutely no reason to use the 32bit
> haskell on OS X (aside from memory usage optimization cases which likely do
> not matter to the *typical* user), and the community should probably update
> the recommendation to reflect this.

The source of the recommendation are the benchmark results presented here:

http://mtnviewmark.wordpress.com/2011/12/07/32-bits-less-is-more/

Note that it's very common to run other GC:ed languages, such as Java
and Python, in 32-bit mode whenever possible. 32-bit almost halves the
memory footprint and thus shortens GC pauses, in particular major
ones, which are O(n) in the size of the heap.

The problem of missing 32-bit C libraries might be a good reason for
us to recommend 64-bit though and leave the 32-bit recommendation to
people who know what they are doing.

Cheers,
Johan
Erik Hesselink | 26 Sep 11:57 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

On Wed, Sep 26, 2012 at 10:58 AM, Johan Tibell <johan.tibell <at> gmail.com> wrote:
> On Wed, Sep 26, 2012 at 7:44 AM, Carter Schonwald
> <carter.schonwald <at> gmail.com> wrote:
>> To the best of my knowledge there is absolutely no reason to use the 32bit
>> haskell on OS X (aside from memory usage optimization cases which likely do
>> not matter to the *typical* user), and the community should probably update
>> the recommendation to reflect this.
>
> The source of the recommendation are the benchmark results presented here:
>
> http://mtnviewmark.wordpress.com/2011/12/07/32-bits-less-is-more/
>
> The problem of missing 32-bit C libraries might be a good reason for
> us to recommend 64-bit though and leave the 32-bit recommendation to
> people who know what they are doing.

We switched to a 64bit GHC recently for this exact reason. The 64bit
libraries are either already installed, or can easily be installed
through e.g. brew. For 32bit libraries, we sometimes had to compile
from source, passing all kinds of flags. The downside for us is
doubling the memory usage, but that's more easily solved (with more
memory). I haven't noticed the performance reduction.

So in short, I think the 64bit version should be the default
recommendation on OS X.

Erik
Johan Tibell | 26 Sep 14:15 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Adding Mark who's the release manager for the platform (and also the
maintainer of the OS X builds).

On Wed, Sep 26, 2012 at 11:57 AM, Erik Hesselink <hesselink <at> gmail.com> wrote:
> On Wed, Sep 26, 2012 at 10:58 AM, Johan Tibell <johan.tibell <at> gmail.com> wrote:
>> On Wed, Sep 26, 2012 at 7:44 AM, Carter Schonwald
>> <carter.schonwald <at> gmail.com> wrote:
>>> To the best of my knowledge there is absolutely no reason to use the 32bit
>>> haskell on OS X (aside from memory usage optimization cases which likely do
>>> not matter to the *typical* user), and the community should probably update
>>> the recommendation to reflect this.
>>
>> The source of the recommendation are the benchmark results presented here:
>>
>> http://mtnviewmark.wordpress.com/2011/12/07/32-bits-less-is-more/
>>
>> The problem of missing 32-bit C libraries might be a good reason for
>> us to recommend 64-bit though and leave the 32-bit recommendation to
>> people who know what they are doing.
>
> We switched to a 64bit GHC recently for this exact reason. The 64bit
> libraries are either already installed, or can easily be installed
> through e.g. brew. For 32bit libraries, we sometimes had to compile
> from source, passing all kinds of flags. The downside for us is
> doubling the memory usage, but that's more easily solved (with more
> memory). I haven't noticed the performance reduction.
>
> So in short, I think the 64bit version should be the default
> recommendation on OS X.
Bob Hutchison | 26 Sep 13:16 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?


On 2012-09-26, at 1:44 AM, Carter Schonwald <carter.schonwald <at> gmail.com> wrote:

what can we (the community ) do to address the fact that the haskell platform installer suggestions for os x are sadly completely backwards? (or am I completely wrong in my personal stance on this matter)


I'd much prefer the 64 bit myself, unfortunately http://hackage.haskell.org/trac/ghc/ticket/7040 affects the current version of the Haskell Platform. It's fixed in 7.6.x. This bug prevents me from using Yesod with the 64 bit version of 7.4.2, and it's not just Yesod affected…

Cheers,
Bob
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Christiaan Baaij | 26 Sep 16:43 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Running gloss [1] programs from GHCi only works with the 32bit version of the latest Haskell Platform.
The 64-bit version just shows a black window and GHCi becomes unresponsive.

I use gloss to display trees and graphs in the functional programming course given at our university.
The ability to work within the interpreter, and able to depict the graphs/trees, really enhances the
student's understanding of their programs.

I am hence glad that there is still the 32-bit recommendation for OS X.

Cheers,

Christiaan

[1] http://hackage.haskell.org/package/gloss - install with: cabal install gloss --flags="-GLUT GLFW"

On Sep 26, 2012, at 1:16 PM, Bob Hutchison wrote:

> 
> On 2012-09-26, at 1:44 AM, Carter Schonwald <carter.schonwald <at> gmail.com> wrote:
> 
>> what can we (the community ) do to address the fact that the haskell platform installer suggestions for os
x are sadly completely backwards? (or am I completely wrong in my personal stance on this matter)
>> 
> 
> I'd much prefer the 64 bit myself, unfortunately http://hackage.haskell.org/trac/ghc/ticket/7040
affects the current version of the Haskell Platform. It's fixed in 7.6.x. This bug prevents me from using
Yesod with the 64 bit version of 7.4.2, and it's not just Yesod affected…
> 
> Cheers,
> Bob
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe
Carter Schonwald | 26 Sep 23:03 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

really? does the 64 bit code work correctly when compiled?

if the compiled version works correctly, could you post a repo of some example codlets that *should work* on ghc 7.6 so i can sort out if its fixable.   There were some similar problems with gtk / cairo for a while on OS X, and i was able to sort out reproducible instructions for getting things to work in that case.





On Wed, Sep 26, 2012 at 10:43 AM, Christiaan Baaij <christiaan.baaij <at> gmail.com> wrote:
Running gloss [1] programs from GHCi only works with the 32bit version of the latest Haskell Platform.
The 64-bit version just shows a black window and GHCi becomes unresponsive.

I use gloss to display trees and graphs in the functional programming course given at our university.
The ability to work within the interpreter, and able to depict the graphs/trees, really enhances the student's understanding of their programs.

I am hence glad that there is still the 32-bit recommendation for OS X.

Cheers,

Christiaan

[1] http://hackage.haskell.org/package/gloss - install with: cabal install gloss --flags="-GLUT GLFW"


On Sep 26, 2012, at 1:16 PM, Bob Hutchison wrote:

>
> On 2012-09-26, at 1:44 AM, Carter Schonwald <carter.schonwald <at> gmail.com> wrote:
>
>> what can we (the community ) do to address the fact that the haskell platform installer suggestions for os x are sadly completely backwards? (or am I completely wrong in my personal stance on this matter)
>>
>
> I'd much prefer the 64 bit myself, unfortunately http://hackage.haskell.org/trac/ghc/ticket/7040 affects the current version of the Haskell Platform. It's fixed in 7.6.x. This bug prevents me from using Yesod with the 64 bit version of 7.4.2, and it's not just Yesod affected…
>
> Cheers,
> Bob
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe


_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Christiaan Baaij | 27 Sep 09:45 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

The behaviour seems to differ between versions of OS X.

A student has OS X 10.8 installed and is observing the described behaviour:
32-bit: interpreted and compiled work correctly
64-bit: only compiled code works correctly

However, I have OS X 10.6, and I'm observing the following behaviour:
32-bit: interpreted and compiled code work correctly
64-bit: interpreted and compiled code work correctly

To test:
cabal install gloss --flags="-GLUT GLFW"
cabal unpack gloss-examples
cd <unpacked-folder>/picture/GameEvent
ghci -fno-ghci-sandbox Main.hs
main

I'll try to find another OS X 10.8 install (I don't want to upgrade) and see if the behaviour still emerges.

On Sep 26, 2012, at 11:03 PM, Carter Schonwald wrote:

> really? does the 64 bit code work correctly when compiled?
> if the compiled version works correctly, could you post a repo of some example codlets that *should work*
on ghc 7.6 so i can sort out if its fixable.   There were some similar problems with gtk / cairo for a while on OS
X, and i was able to sort out reproducible instructions for getting things to work in that case.
Lyndon Maydwell | 27 Sep 09:49 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

I'm experiencing the same issues with compiled 64 bit working
correctly, but interpreted causing all sorts of issues with Scotty.

On Thu, Sep 27, 2012 at 3:45 PM, Christiaan Baaij
<christiaan.baaij <at> gmail.com> wrote:
> The behaviour seems to differ between versions of OS X.
>
> A student has OS X 10.8 installed and is observing the described behaviour:
> 32-bit: interpreted and compiled work correctly
> 64-bit: only compiled code works correctly
>
> However, I have OS X 10.6, and I'm observing the following behaviour:
> 32-bit: interpreted and compiled code work correctly
> 64-bit: interpreted and compiled code work correctly
>
> To test:
> cabal install gloss --flags="-GLUT GLFW"
> cabal unpack gloss-examples
> cd <unpacked-folder>/picture/GameEvent
> ghci -fno-ghci-sandbox Main.hs
> main
>
> I'll try to find another OS X 10.8 install (I don't want to upgrade) and see if the behaviour still emerges.
>
>
> On Sep 26, 2012, at 11:03 PM, Carter Schonwald wrote:
>
>> really? does the 64 bit code work correctly when compiled?
>> if the compiled version works correctly, could you post a repo of some example codlets that *should work*
on ghc 7.6 so i can sort out if its fixable.   There were some similar problems with gtk / cairo for a while on OS
X, and i was able to sort out reproducible instructions for getting things to work in that case.
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe
Carter Schonwald | 28 Sep 07:56 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

do these problems also happen if your'e using the glut backend? (because if its only glfw that has problems, then its something wrong in the ffi code, 

but if its both, that suggests there may be some sort of systematic problem?)

<at> Lyndon, that sounds like a bug... especially since scotty seems to have no C code in package... have you filed a bug report with the maintainers thats reproducible?

On Thu, Sep 27, 2012 at 3:49 AM, Lyndon Maydwell <maydwell <at> gmail.com> wrote:
I'm experiencing the same issues with compiled 64 bit working
correctly, but interpreted causing all sorts of issues with Scotty.

On Thu, Sep 27, 2012 at 3:45 PM, Christiaan Baaij
<christiaan.baaij <at> gmail.com> wrote:
> The behaviour seems to differ between versions of OS X.
>
> A student has OS X 10.8 installed and is observing the described behaviour:
> 32-bit: interpreted and compiled work correctly
> 64-bit: only compiled code works correctly
>
> However, I have OS X 10.6, and I'm observing the following behaviour:
> 32-bit: interpreted and compiled code work correctly
> 64-bit: interpreted and compiled code work correctly
>
> To test:
> cabal install gloss --flags="-GLUT GLFW"
> cabal unpack gloss-examples
> cd <unpacked-folder>/picture/GameEvent
> ghci -fno-ghci-sandbox Main.hs
> main
>
> I'll try to find another OS X 10.8 install (I don't want to upgrade) and see if the behaviour still emerges.
>
>
> On Sep 26, 2012, at 11:03 PM, Carter Schonwald wrote:
>
>> really? does the 64 bit code work correctly when compiled?
>> if the compiled version works correctly, could you post a repo of some example codlets that *should work* on ghc 7.6 so i can sort out if its fixable.   There were some similar problems with gtk / cairo for a while on OS X, and i was able to sort out reproducible instructions for getting things to work in that case.
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Lyndon Maydwell | 28 Sep 08:06 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Carter:

Not yet.

I'll get round to it once I'm done with with an Agda presentation I'm
working on. Until then I can't afford to break my environment.

On Fri, Sep 28, 2012 at 1:56 PM, Carter Schonwald
<carter.schonwald <at> gmail.com> wrote:
> do these problems also happen if your'e using the glut backend? (because if
> its only glfw that has problems, then its something wrong in the ffi code,
> but if its both, that suggests there may be some sort of systematic
> problem?)
>
>  <at> Lyndon, that sounds like a bug... especially since scotty seems to have no
> C code in package... have you filed a bug report with the maintainers thats
> reproducible?
>
>
> On Thu, Sep 27, 2012 at 3:49 AM, Lyndon Maydwell <maydwell <at> gmail.com> wrote:
>>
>> I'm experiencing the same issues with compiled 64 bit working
>> correctly, but interpreted causing all sorts of issues with Scotty.
>>
>> On Thu, Sep 27, 2012 at 3:45 PM, Christiaan Baaij
>> <christiaan.baaij <at> gmail.com> wrote:
>> > The behaviour seems to differ between versions of OS X.
>> >
>> > A student has OS X 10.8 installed and is observing the described
>> > behaviour:
>> > 32-bit: interpreted and compiled work correctly
>> > 64-bit: only compiled code works correctly
>> >
>> > However, I have OS X 10.6, and I'm observing the following behaviour:
>> > 32-bit: interpreted and compiled code work correctly
>> > 64-bit: interpreted and compiled code work correctly
>> >
>> > To test:
>> > cabal install gloss --flags="-GLUT GLFW"
>> > cabal unpack gloss-examples
>> > cd <unpacked-folder>/picture/GameEvent
>> > ghci -fno-ghci-sandbox Main.hs
>> > main
>> >
>> > I'll try to find another OS X 10.8 install (I don't want to upgrade) and
>> > see if the behaviour still emerges.
>> >
>> >
>> > On Sep 26, 2012, at 11:03 PM, Carter Schonwald wrote:
>> >
>> >> really? does the 64 bit code work correctly when compiled?
>> >> if the compiled version works correctly, could you post a repo of some
>> >> example codlets that *should work* on ghc 7.6 so i can sort out if its
>> >> fixable.   There were some similar problems with gtk / cairo for a while on
>> >> OS X, and i was able to sort out reproducible instructions for getting
>> >> things to work in that case.
>> >
>> > _______________________________________________
>> > Haskell-Cafe mailing list
>> > Haskell-Cafe <at> haskell.org
>> > http://www.haskell.org/mailman/listinfo/haskell-cafe
>
>
Christiaan Baaij | 28 Sep 13:16 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

The GLUT-backend calls system.exit when the window is closed, because 'exitMainLoop' is only defined
within freeglut, which is not by default installed on non-linux platforms.
There is hence no real point in running gloss applications with the GLUT-backend from GHCi.

I'll try to find a mountain lion install, and test if there's a difference between GLUT and GLFW 

On Sep 28, 2012, at 7:56 AM, Carter Schonwald wrote:

> do these problems also happen if your'e using the glut backend? (because if its only glfw that has
problems, then its something wrong in the ffi code, 
> but if its both, that suggests there may be some sort of systematic problem?)
> 
>  <at> Lyndon, that sounds like a bug... especially since scotty seems to have no C code in package... have you
filed a bug report with the maintainers thats reproducible?
Christiaan Baaij | 8 Oct 12:28 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Hi,

I finally found another OS X mountain lion install and can confirm the behaviour I described earlier:
32-bit: compiled code works, interpreted code works
64-bit: compiled code works, interpreted code fails

Here's the test case:
- cabal install gloss --flags"-GLUT GLFW"
- cabal unpack gloss-examples
- cd gloss-examples-1.7.6.2/picture/GameEvent
- ghci -fno-ghci-sandbox Main.hs
- main

I get the following crash report: http://pastebin.com/jZjfFtm7

The weird thing is the following:
When I run 'ghci' from inside 'gdb' (to find the origin for the segfault), everything works fine:
ghci: segfault
ghci from gdb: everything works

I have no idea what's going on, so if anyone has any pointers on how to make sure ghci behaves the same in gdb
please let me know.

Cheers,

Christiaan

On Sep 28, 2012, at 1:16 PM, Christiaan Baaij wrote:

> The GLUT-backend calls system.exit when the window is closed, because 'exitMainLoop' is only defined
within freeglut, which is not by default installed on non-linux platforms.
> There is hence no real point in running gloss applications with the GLUT-backend from GHCi.
> 
> I'll try to find a mountain lion install, and test if there's a difference between GLUT and GLFW 
> 
> On Sep 28, 2012, at 7:56 AM, Carter Schonwald wrote:
> 
>> do these problems also happen if your'e using the glut backend? (because if its only glfw that has
problems, then its something wrong in the ffi code, 
>> but if its both, that suggests there may be some sort of systematic problem?)
>> 
>>  <at> Lyndon, that sounds like a bug... especially since scotty seems to have no C code in package... have you
filed a bug report with the maintainers thats reproducible?
> 
brandon s allbery kf8nh | 8 Oct 16:00 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

On Monday, 8 October 2012 at 06:28, Christiaan Baaij wrote:
ghci: segfault
ghci from gdb: everything works
This makes me suspect something that gets disabled when debugging, such as address space randomization and the like.  I did not think ML handled that any differently from Lion, though.

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b <at> gmail.com                                  ballbery <at> sinenomine.net
unix/linux, openafs, kerberos, infrastructure          http://sinenomine.net

Sent with Sparrow

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Johan Tibell | 8 Oct 16:25 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

On Mon, Oct 8, 2012 at 3:28 AM, Christiaan Baaij
<christiaan.baaij <at> gmail.com> wrote:
> Hi,
>
> I finally found another OS X mountain lion install and can confirm the behaviour I described earlier:
> 32-bit: compiled code works, interpreted code works
> 64-bit: compiled code works, interpreted code fails
>
> Here's the test case:
> - cabal install gloss --flags"-GLUT GLFW"
> - cabal unpack gloss-examples
> - cd gloss-examples-1.7.6.2/picture/GameEvent
> - ghci -fno-ghci-sandbox Main.hs
> - main
>
> I get the following crash report: http://pastebin.com/jZjfFtm7
>
> The weird thing is the following:
> When I run 'ghci' from inside 'gdb' (to find the origin for the segfault), everything works fine:
> ghci: segfault
> ghci from gdb: everything works
>
> I have no idea what's going on, so if anyone has any pointers on how to make sure ghci behaves the same in gdb
please let me know.

Could you please file a bug report at:

http://hackage.haskell.org/trac/ghc/

Thanks!
Eugene Kirpichov | 8 Oct 21:08 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Johan, should I also file the bugreport "remove the suggestion to
install 32-bit platform" there, or is there a different place for bugs
of the platform website?

On Mon, Oct 8, 2012 at 7:25 AM, Johan Tibell <johan.tibell <at> gmail.com> wrote:
> On Mon, Oct 8, 2012 at 3:28 AM, Christiaan Baaij
> <christiaan.baaij <at> gmail.com> wrote:
>> Hi,
>>
>> I finally found another OS X mountain lion install and can confirm the behaviour I described earlier:
>> 32-bit: compiled code works, interpreted code works
>> 64-bit: compiled code works, interpreted code fails
>>
>> Here's the test case:
>> - cabal install gloss --flags"-GLUT GLFW"
>> - cabal unpack gloss-examples
>> - cd gloss-examples-1.7.6.2/picture/GameEvent
>> - ghci -fno-ghci-sandbox Main.hs
>> - main
>>
>> I get the following crash report: http://pastebin.com/jZjfFtm7
>>
>> The weird thing is the following:
>> When I run 'ghci' from inside 'gdb' (to find the origin for the segfault), everything works fine:
>> ghci: segfault
>> ghci from gdb: everything works
>>
>> I have no idea what's going on, so if anyone has any pointers on how to make sure ghci behaves the same in gdb
please let me know.
>
> Could you please file a bug report at:
>
> http://hackage.haskell.org/trac/ghc/
>
> Thanks!
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe

--

-- 
Eugene Kirpichov
http://www.linkedin.com/in/eugenekirpichov
We're hiring! http://tinyurl.com/mirantis-openstack-engineer
Carter Schonwald | 8 Oct 22:06 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Eugene: I think thats a bug / ticket for the haskell platform trac, rather than ghc's trac.


look forward to seeing how to reproduce those problems / helping fix em! 

On Mon, Oct 8, 2012 at 3:08 PM, Eugene Kirpichov <ekirpichov <at> gmail.com> wrote:
Johan, should I also file the bugreport "remove the suggestion to
install 32-bit platform" there, or is there a different place for bugs
of the platform website?

On Mon, Oct 8, 2012 at 7:25 AM, Johan Tibell <johan.tibell <at> gmail.com> wrote:
> On Mon, Oct 8, 2012 at 3:28 AM, Christiaan Baaij
> <christiaan.baaij <at> gmail.com> wrote:
>> Hi,
>>
>> I finally found another OS X mountain lion install and can confirm the behaviour I described earlier:
>> 32-bit: compiled code works, interpreted code works
>> 64-bit: compiled code works, interpreted code fails
>>
>> Here's the test case:
>> - cabal install gloss --flags"-GLUT GLFW"
>> - cabal unpack gloss-examples
>> - cd gloss-examples-1.7.6.2/picture/GameEvent
>> - ghci -fno-ghci-sandbox Main.hs
>> - main
>>
>> I get the following crash report: http://pastebin.com/jZjfFtm7
>>
>> The weird thing is the following:
>> When I run 'ghci' from inside 'gdb' (to find the origin for the segfault), everything works fine:
>> ghci: segfault
>> ghci from gdb: everything works
>>
>> I have no idea what's going on, so if anyone has any pointers on how to make sure ghci behaves the same in gdb please let me know.
>
> Could you please file a bug report at:
>
> http://hackage.haskell.org/trac/ghc/
>
> Thanks!
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe



_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Mark Lentczner | 9 Oct 06:52 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

I'm the source of the 32-bit recommendation, and the HP Mac distribution builder


To summarize what I read in this thread:
  1. 32-bit GHC/HP didn't work with 64-bit Cario libs
  2. Some libs available via brew were 64-bit, and 32-bit ones would have to be compiled
  3. There is still some bug with 64-bit ghci and some graphics libs
  4. There is a ghc bug with 64-bit on mac (bug #7040), which isn't fixed until 7.6
There seemed to be the implication that a 64-bit ghc would work with 32-bit libs, but I don't think that's true. Mac doesn't (generally) support mixed modes in the same executable. All system libs are shipped dual-architecture. I don't think there are any pre-installed libs that are shipped 64-bit only. The problem seen with Cairo would cut both ways: If one had installed the 32-bit version of Cairo, one would see the same problem with the 64-bit HP: "wrong architecture".

Since code compiled with the 32-bit system is both faster, and uses less memory, and it has been the case that all libs are either shipped dual-arch, or easily available as 32-bit, and there were known problems with the 64-bit version for some use cases, it seemed to me to be best to suggest the 32-bit version by default.

The major source of the problems in the OP, seem to be that MacPorts and/or brew don't appear to follow the Mac OS X lib standard of installing libs dual arch. A brief look at the MacPorts page indicated that there were various rules (OS version and processor version) that determined which arch. it built by default. Perhaps we should tell people to install the HP architecture that matches the architecture that they use for MacPorts or brew. However, I bet most people don't know, so we'd need a pointer to where they could find out the defaults for those systems, or how to establish what their system is using.

Finally, I note that HP 2012.4.0.0 (out in a month) will ship with GHC 7.4.2, and so will still have the above bugs.

- Mark

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Eugene Kirpichov | 11 Oct 19:41 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

Hi Mark,

Thank you for the clarification.

I think, then, that we should indeed provide a link to a notice about
native libraries - like, "if you're using native libraries, make sure
their architecture is 32-bit  or universal as well; in case of
MacPorts you can achieve this by doing port install libfoo +universal;
in case of brew by brew install libfoo --universal".

On Mon, Oct 8, 2012 at 9:52 PM, Mark Lentczner <mark.lentczner <at> gmail.com> wrote:
> I'm the source of the 32-bit recommendation, and the HP Mac distribution
> builder
>
> To summarize what I read in this thread:
>
> 32-bit GHC/HP didn't work with 64-bit Cario libs
> Some libs available via brew were 64-bit, and 32-bit ones would have to be
> compiled
> There is still some bug with 64-bit ghci and some graphics libs
> There is a ghc bug with 64-bit on mac (bug #7040), which isn't fixed until
> 7.6
>
> There seemed to be the implication that a 64-bit ghc would work with 32-bit
> libs, but I don't think that's true. Mac doesn't (generally) support mixed
> modes in the same executable. All system libs are shipped dual-architecture.
> I don't think there are any pre-installed libs that are shipped 64-bit only.
> The problem seen with Cairo would cut both ways: If one had installed the
> 32-bit version of Cairo, one would see the same problem with the 64-bit HP:
> "wrong architecture".
>
> Since code compiled with the 32-bit system is both faster, and uses less
> memory, and it has been the case that all libs are either shipped dual-arch,
> or easily available as 32-bit, and there were known problems with the 64-bit
> version for some use cases, it seemed to me to be best to suggest the 32-bit
> version by default.
>
> The major source of the problems in the OP, seem to be that MacPorts and/or
> brew don't appear to follow the Mac OS X lib standard of installing libs
> dual arch. A brief look at the MacPorts page indicated that there were
> various rules (OS version and processor version) that determined which arch.
> it built by default. Perhaps we should tell people to install the HP
> architecture that matches the architecture that they use for MacPorts or
> brew. However, I bet most people don't know, so we'd need a pointer to where
> they could find out the defaults for those systems, or how to establish what
> their system is using.
>
> Finally, I note that HP 2012.4.0.0 (out in a month) will ship with GHC
> 7.4.2, and so will still have the above bugs.
>
> - Mark
>
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe <at> haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

--

-- 
Eugene Kirpichov
http://www.linkedin.com/in/eugenekirpichov
We're hiring! http://tinyurl.com/mirantis-openstack-engineer
Christiaan Baaij | 9 Oct 12:18 2012
Picon

Re: 64-bit vs 32-bit haskell platform on Mac: misleading notice on Platform website?

I've filed a bug report: http://hackage.haskell.org/trac/ghc/ticket/7314

My only problem is that I don't have OS X 10.8 installed on my machine.
I'm running OS X 10.6, and am not inclined to upgrade.
So providing feedback on the questions related to the bug report will be hard and/or take some time.

-- Christiaan

On Oct 8, 2012, at 4:25 PM, Johan Tibell wrote:

> Could you please file a bug report at:
> 
> http://hackage.haskell.org/trac/ghc/
> 
> Thanks!

Gmane