On Thursday, August 1, 2013 8:39:12 AM UTC+2, degski wrote:
>
> "Though it seems that building MPIR under VS is harder than on Linux!"
>
> I'll try and help you out.
>
> 0. Install VS.
> 1. Install yasm following the instructions on the yasm home-page.
> 3. Goto MPIR-homepage and download tar-ball.
> 4. Get 7zip, install and extract tar-ball in convenient location.
> 5. Fire-up VS.
> 6. Open sln-file in build folder of MPIR-source.
> 7. Select config you would like, debug/release, Win32, x64.
> 8. Build, wait 1 minute.
> 9. Enjoy.
>
> It hardly can get any easier than that.
>
> No harm meant, I was just kidding "you have to download two files rather 
than one". 

>
>
> On 1 August 2013 00:45, Jean-Pierre Flori <[email protected] 
> <javascript:>>wrote:
>
>>
>>
>> On Wednesday, July 31, 2013 11:26:04 PM UTC+2, Bill Hart wrote:
>>
>>>
>>>
>>>
>>> On 31 July 2013 18:24, Jean-Pierre Flori <[email protected]> wrote:
>>>
>>>> Dear all,
>>>>
>>>> I'm currently working on some pieces of the build system.
>>>> I must admit I don't really like the fact that MPIR ships a modified 
>>>> yasm source tree.
>>>> I think it would be cleaner, and easier to maintain, to ship a tarball 
>>>> and untar it, or to the least a vanilla source tree, possibly patch it and 
>>>>  
>>>> then configure/build it if needed (i already added options to use a 
>>>> system-wide or user-provided yasm, when autotools are used at least).
>>>>
>>>
>>> This should be possible. As far as I know the Windows MSVC build uses a 
>>> precompiled yasm, not the one in the tree, so everything should be fine if 
>>> yasm is built separately.
>>>
>>>  The downside is that GMP builds out-of-the-box without requiring one 
>>> to download a separate assembler. I would abandon yasm for *nix if it were 
>>> not for the fact that it is much easier for Brian Gladman to convert *nix 
>>> assembler to Windows format if it is already in Yasm format (Intel format 
>>> assembler is much more popular that AT&T format on Windows, I believe).
>>>
>> I didnot mean to not ship Yasm anymore, rather to ship a vanilla one, 
>> either through a tarball or through a source tree (as of now, but without 
>> modification if possible).
>> Then updating Yasm would be trivial (ok, you might have to rename a 
>> tarball/directory, or modify a string somewhere).
>>  
>>
>>>  
>>>
>>>> A little like what we do in Sage when we patch upstream software (like 
>>>> MPIR!).
>>>>
>>>> Do you really need to pass specific options to yasm? or a "generic" 
>>>> (./configure && make) build would do?
>>>>
>>>
>>> I believe some options are necessarily passed to Yasm.
>>>
>> With the current setup everything is passed verbatim ... whence the need 
>> to modify yasm build system, because the vanilla build system will error 
>> out if unrecognized option (let's say --enable-gmpcompat) are passed.
>>
>> It's possible to make it so that the top level configure option are 
>> discarded when going into a subdir.
>> It's more of an hack than a real option of autotools but should avoid to 
>> modify yasm build system.
>> That's what I had in mind and why I was asking if there were situation 
>> where option passed to the main configure script were really relevant to 
>> yasm (from what I've seen it's not the case, on the contrary we have to 
>> modify yasm so that it lives with MPIR superfluous options).
>>
>>  
>>>
>>>> That's important because the main modification I see is to let the 
>>>> autotools stuff recognize options given to MPIr and automatically passed 
>>>> to 
>>>> yasm.
>>>>
>>>
>>> I think the options are not the same as the ones passed to MPIR. But it 
>>> is such a long time since I worked on this, I don't recall the details.
>>>  
>>>
>>>>
>>>> My other question is about the VS builds because I never tried them and 
>>>> feel completely incompetent.
>>>> With VS, is yasm always built? 
>>>
>>>
>>> As far as I recall, the user downloads a special version of Yasm 
>>> maintained by Peter Tortall, installs that in a certain location on their 
>>> Windows, then builds MPIR in Visual Studio.
>>>
>>> That would be good news for me!
>> (Though it seems that building MPIR under VS is harder than on Linux!)
>>  
>>
>>> Bill.
>>>
>>> Is it possible to let VS untar something? or does yasm directory have to 
>>>> be uncompressed?
>>>>
>>>> Best,
>>>> JP
>>>>
>>>>  -- 
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "mpir-devel" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to mpir-devel+...@**googlegroups.com.
>>>> To post to this group, send email to [email protected].
>>>>
>>>> Visit this group at 
>>>> http://groups.google.com/**group/mpir-devel<http://groups.google.com/group/mpir-devel>
>>>> .
>>>> For more options, visit 
>>>> https://groups.google.com/**groups/opt_out<https://groups.google.com/groups/opt_out>
>>>> .
>>>>  
>>>>  
>>>>
>>>
>>>  -- 
>> You received this message because you are subscribed to the Google Groups 
>> "mpir-devel" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected] <javascript:>.
>> To post to this group, send email to [email protected]<javascript:>
>> .
>> Visit this group at http://groups.google.com/group/mpir-devel.
>> For more options, visit https://groups.google.com/groups/opt_out.
>>  
>>  
>>
>
>
>
> -- 
> Sign the petition: 
> https://optin.stopwatching.us/<https://optin.stopwatching.us/>
>  

-- 
You received this message because you are subscribed to the Google Groups 
"mpir-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/mpir-devel.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to