Quantcast

ghc-7.4.1-rc1 recompilation checking

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate
star

ghc-7.4.1-rc1 recompilation checking

Christian Maeder-2
Hi,

full recompilation with ghc-7.4.1-rc1 is already triggered by a changed
"-o" option, which is inconvenient when creating different binaries with
shared modules. (see below)

It is no problem if I always omit the -o option and get binaries named
like my input file, though.

Can/should this behaviour of ghc-7.4 be relaxed?

Cheers Christian

-bash-3.00$ ghc B -o b
[1 of 2] Compiling C                ( C.hs, C.o )
[2 of 2] Compiling Main             ( B.hs, B.o )
Linking b ...
-bash-3.00$ ghc A -o a
[1 of 2] Compiling C                ( C.hs, C.o )
[2 of 2] Compiling Main             ( A.hs, A.o )
Linking a ...

_______________________________________________
Glasgow-haskell-users mailing list
[hidden email]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate
star

Re: ghc-7.4.1-rc1 recompilation checking

Simon Marlow-7
On 27/12/2011 13:07, Christian Maeder wrote:

> Hi,
>
> full recompilation with ghc-7.4.1-rc1 is already triggered by a changed
> "-o" option, which is inconvenient when creating different binaries with
> shared modules. (see below)
>
> It is no problem if I always omit the -o option and get binaries named
> like my input file, though.
>
> Can/should this behaviour of ghc-7.4 be relaxed?
>
> Cheers Christian
>
> -bash-3.00$ ghc B -o b
> [1 of 2] Compiling C ( C.hs, C.o )
> [2 of 2] Compiling Main ( B.hs, B.o )
> Linking b ...
> -bash-3.00$ ghc A -o a
> [1 of 2] Compiling C ( C.hs, C.o )
> [2 of 2] Compiling Main ( A.hs, A.o )
> Linking a ...

Good catch - I'll get a fix for this into 7.4.1.

Cheers,
        Simon


_______________________________________________
Glasgow-haskell-users mailing list
[hidden email]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate
star

Re: ghc-7.4.1-rc1 recompilation checking

Christian Maeder-2
This seems not to be fixed in Release Candidate 2

Cheers Christian

Am 03.01.2012 11:29, schrieb Simon Marlow:

> On 27/12/2011 13:07, Christian Maeder wrote:
>> Hi,
>>
>> full recompilation with ghc-7.4.1-rc1 is already triggered by a changed
>> "-o" option, which is inconvenient when creating different binaries with
>> shared modules. (see below)
>>
>> It is no problem if I always omit the -o option and get binaries named
>> like my input file, though.
>>
>> Can/should this behaviour of ghc-7.4 be relaxed?
>>
>> Cheers Christian
>>
>> -bash-3.00$ ghc B -o b
>> [1 of 2] Compiling C ( C.hs, C.o )
>> [2 of 2] Compiling Main ( B.hs, B.o )
>> Linking b ...
>> -bash-3.00$ ghc A -o a
>> [1 of 2] Compiling C ( C.hs, C.o )
>> [2 of 2] Compiling Main ( A.hs, A.o )
>> Linking a ...
>
> Good catch - I'll get a fix for this into 7.4.1.
>
> Cheers,
> Simon
>
>

_______________________________________________
Glasgow-haskell-users mailing list
[hidden email]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate
star

Re: ghc-7.4.1-rc1 recompilation checking

Simon Marlow-7
On 30/01/2012 09:45, Christian Maeder wrote:
> This seems not to be fixed in Release Candidate 2

Thanks for letting us know.  Ian, please merge this patch:

commit f283c39119b333e4249420fa96b93652e0fbaec1
Author: Simon Marlow <[hidden email]>
Date:   Tue Jan 3 10:28:39 2012 +0000

     Don't record outputFile in the hashed flags

     We don't want

         ghc --make M -o <file>

     to force recompilation of all modules when <file> changes.  The -o
     value is already taken into account by the recompilation machinery
     when we check the modification time on the object file or the
     executable.


Cheers,
        Simon



> Cheers Christian
>
> Am 03.01.2012 11:29, schrieb Simon Marlow:
>> On 27/12/2011 13:07, Christian Maeder wrote:
>>> Hi,
>>>
>>> full recompilation with ghc-7.4.1-rc1 is already triggered by a changed
>>> "-o" option, which is inconvenient when creating different binaries with
>>> shared modules. (see below)
>>>
>>> It is no problem if I always omit the -o option and get binaries named
>>> like my input file, though.
>>>
>>> Can/should this behaviour of ghc-7.4 be relaxed?
>>>
>>> Cheers Christian
>>>
>>> -bash-3.00$ ghc B -o b
>>> [1 of 2] Compiling C ( C.hs, C.o )
>>> [2 of 2] Compiling Main ( B.hs, B.o )
>>> Linking b ...
>>> -bash-3.00$ ghc A -o a
>>> [1 of 2] Compiling C ( C.hs, C.o )
>>> [2 of 2] Compiling Main ( A.hs, A.o )
>>> Linking a ...
>>
>> Good catch - I'll get a fix for this into 7.4.1.
>>
>> Cheers,
>> Simon
>>
>>


_______________________________________________
Glasgow-haskell-users mailing list
[hidden email]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate
star

Re: ghc-7.4.1-rc1 recompilation checking

Ian Lynagh
On Mon, Jan 30, 2012 at 10:02:42AM +0000, Simon Marlow wrote:

> On 30/01/2012 09:45, Christian Maeder wrote:
> >This seems not to be fixed in Release Candidate 2
>
> Thanks for letting us know.  Ian, please merge this patch:
>
> commit f283c39119b333e4249420fa96b93652e0fbaec1
> Author: Simon Marlow <[hidden email]>
> Date:   Tue Jan 3 10:28:39 2012 +0000
>
>     Don't record outputFile in the hashed flags

Merged as 5db3b4cff15fbcaa74c881b5dd5e39f7cade08d5.


Thanks
Ian


_______________________________________________
Glasgow-haskell-users mailing list
[hidden email]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Loading...