libffi

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
12 messages Options
Reply | Threaded
Open this post in threaded view
|

libffi

Simon Peyton Jones
I'm trying to build HEAD.  I get a problem with libffi.  Not surprising because in ghc-tarballs I see

git log -3
commit f190b3ce329422e13cbe1b5dad030058ca4bdda7
Author: Ian Lynagh <ian at well-typed.com>
Date:   Tue Jul 30 15:56:23 2013 +0100

    Remove libffi tarball

    It's now in its own repo

But ./sync-all get doesn't get a libffi repo.  Nor is it listed in ./packages.

Does anyone have any idea where this new libffi repo is, and why it isn't listed?

Thanks

Simon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20130803/cc3f35ed/attachment.htm>

Reply | Threaded
Open this post in threaded view
|

libffi

Erik de Castro Lopo-34
Simon Peyton-Jones wrote:

> But ./sync-all get doesn't get a libffi repo.  Nor is it listed in ./packages.
>
> Does anyone have any idea where this new libffi repo is, and why it isn't listed?

Simon,

I think you need to do a 'git pull' first to get the latest version of the
sync-all script and when you run that you should get a libffi-tarballs
directory.

Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/



Reply | Threaded
Open this post in threaded view
|

libffi

Simon Peyton Jones
|  I think you need to do a 'git pull' first to get the latest version of the
|  sync-all script and when you run that you should get a libffi-tarballs

I'd done that (repeatedly).  BUT, as has often happened before, I'd forgotten that my tree was on a different branch, and pulling of course has no effect on the branch.   Rats.   Note to self: whenever anything odd happens, check you are on 'master'.

Sorry for the noise

Simon

|  -----Original Message-----
|  From: ghc-devs [mailto:ghc-devs-bounces at haskell.org] On Behalf Of Erik de
|  Castro Lopo
|  Sent: 03 August 2013 08:25
|  To: ghc-devs at haskell.org
|  Subject: Re: libffi
|  
|  Simon Peyton-Jones wrote:
|  
|  > But ./sync-all get doesn't get a libffi repo.  Nor is it listed in ./packages.
|  >
|  > Does anyone have any idea where this new libffi repo is, and why it isn't listed?
|  
|  Simon,
|  
|  I think you need to do a 'git pull' first to get the latest version of the
|  sync-all script and when you run that you should get a libffi-tarballs
|  directory.
|  
|  Cheers,
|  Erik
|  --
|  ----------------------------------------------------------------------
|  Erik de Castro Lopo
|  http://www.mega-nerd.com/
|  
|  _______________________________________________
|  ghc-devs mailing list
|  ghc-devs at haskell.org
|  http://www.haskell.org/mailman/listinfo/ghc-devs





Reply | Threaded
Open this post in threaded view
|

Multiple repositories, submodules, subtrees and whatnot (Was: Re: libffi)

Nicolas Trangez
Given the recent debate about using submodules or subtrees, and the
occasional issues with sync-all: did anyone ever look into using the
'repo' tool? Created by Google for Android development (AFAIK), but also
used by several other projects. See https://code.google.com/p/git-repo/

Nicolas

On Sat, 2013-08-03 at 15:41 +0000, Simon Peyton-Jones wrote:

> |  I think you need to do a 'git pull' first to get the latest version of the
> |  sync-all script and when you run that you should get a libffi-tarballs
>
> I'd done that (repeatedly).  BUT, as has often happened before, I'd forgotten that my tree was on a different branch, and pulling of course has no effect on the branch.   Rats.   Note to self: whenever anything odd happens, check you are on 'master'.
>
> Sorry for the noise
>
> Simon
>
> |  -----Original Message-----
> |  From: ghc-devs [mailto:ghc-devs-bounces at haskell.org] On Behalf Of Erik de
> |  Castro Lopo
> |  Sent: 03 August 2013 08:25
> |  To: ghc-devs at haskell.org
> |  Subject: Re: libffi
> |  
> |  Simon Peyton-Jones wrote:
> |  
> |  > But ./sync-all get doesn't get a libffi repo.  Nor is it listed in ./packages.
> |  >
> |  > Does anyone have any idea where this new libffi repo is, and why it isn't listed?
> |  
> |  Simon,
> |  
> |  I think you need to do a 'git pull' first to get the latest version of the
> |  sync-all script and when you run that you should get a libffi-tarballs
> |  directory.
> |  
> |  Cheers,
> |  Erik
> |  --
> |  ----------------------------------------------------------------------
> |  Erik de Castro Lopo
> |  http://www.mega-nerd.com/
> |  
> |  _______________________________________________
> |  ghc-devs mailing list
> |  ghc-devs at haskell.org
> |  http://www.haskell.org/mailman/listinfo/ghc-devs
>
>
>
> _______________________________________________
> ghc-devs mailing list
> ghc-devs at haskell.org
> http://www.haskell.org/mailman/listinfo/ghc-devs





Reply | Threaded
Open this post in threaded view
|

libffi

Erik de Castro Lopo-34
In reply to this post by Simon Peyton Jones
Simon Peyton-Jones wrote:

> |  I think you need to do a 'git pull' first to get the latest version of the
> |  sync-all script and when you run that you should get a libffi-tarballs
>
> I'd done that (repeatedly).  BUT, as has often happened before, I'd forgotten
> that my tree was on a different branch, and pulling of course has no effect on
> the branch.   Rats.   Note to self: whenever anything odd happens, check you
> are on 'master'.

Better yet, the sync-all script should detect whether it is on master and
error out (with a decent error message) if not.

That way you (and I and others) don't have to remember this particular failure
mode.

My Perl coding skills are rather rusty, but I'll have a go at this.

Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/



Reply | Threaded
Open this post in threaded view
|

libffi

Erik de Castro Lopo-34
Erik de Castro Lopo wrote:

> Simon Peyton-Jones wrote:
>
> > I'd done that (repeatedly).  BUT, as has often happened before, I'd forgotten
> > that my tree was on a different branch, and pulling of course has no effect on
> > the branch.   Rats.   Note to self: whenever anything odd happens, check you
> > are on 'master'.
>
> Better yet, the sync-all script should detect whether it is on master and
> error out (with a decent error message) if not.
>
> That way you (and I and others) don't have to remember this particular failure
> mode.
>
> My Perl coding skills are rather rusty, but I'll have a go at this.

Attached is a patch that checks that the current branch is master
and bails if it isn't. If no one sees anything obviously wrong
with this I'll push it.

Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-sync-all-Bail-out-if-not-on-master-branch.patch
Type: text/x-diff
Size: 937 bytes
Desc: not available
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20130804/216de6ab/attachment.patch>

Reply | Threaded
Open this post in threaded view
|

libffi

Simon Peyton Jones
Sounds good to me -- thank you!  Others: do take a look if you can.

Simon

|  -----Original Message-----
|  From: ghc-devs [mailto:ghc-devs-bounces at haskell.org] On Behalf Of Erik de
|  Castro Lopo
|  Sent: 04 August 2013 02:59
|  To: ghc-devs at haskell.org
|  Subject: Re: libffi
|  
|  Erik de Castro Lopo wrote:
|  
|  > Simon Peyton-Jones wrote:
|  >
|  > > I'd done that (repeatedly).  BUT, as has often happened before, I'd
|  > > forgotten that my tree was on a different branch, and pulling of course has no
|  effect on
|  > > the branch.   Rats.   Note to self: whenever anything odd happens, check you
|  > > are on 'master'.
|  >
|  > Better yet, the sync-all script should detect whether it is on master
|  > and error out (with a decent error message) if not.
|  >
|  > That way you (and I and others) don't have to remember this particular
|  > failure mode.
|  >
|  > My Perl coding skills are rather rusty, but I'll have a go at this.
|  
|  Attached is a patch that checks that the current branch is master and bails if it
|  isn't. If no one sees anything obviously wrong with this I'll push it.
|  
|  Cheers,
|  Erik
|  --
|  ----------------------------------------------------------------------
|  Erik de Castro Lopo
|  http://www.mega-nerd.com/




Reply | Threaded
Open this post in threaded view
|

libffi

Patrick Palka
In reply to this post by Erik de Castro Lopo-34
I think the branch check should only be done when performing a pull
command. That way one can still do ./sync-all fetch, ./sync-all diff, etc
when on a separate branch.


On Sat, Aug 3, 2013 at 9:59 PM, Erik de Castro Lopo <mle+hs at mega-nerd.com>wrote:

> Erik de Castro Lopo wrote:
>
> > Simon Peyton-Jones wrote:
> >
> > > I'd done that (repeatedly).  BUT, as has often happened before, I'd
> forgotten
> > > that my tree was on a different branch, and pulling of course has no
> effect on
> > > the branch.   Rats.   Note to self: whenever anything odd happens,
> check you
> > > are on 'master'.
> >
> > Better yet, the sync-all script should detect whether it is on master and
> > error out (with a decent error message) if not.
> >
> > That way you (and I and others) don't have to remember this particular
> failure
> > mode.
> >
> > My Perl coding skills are rather rusty, but I'll have a go at this.
>
> Attached is a patch that checks that the current branch is master
> and bails if it isn't. If no one sees anything obviously wrong
> with this I'll push it.
>
> Cheers,
> Erik
> --
> ----------------------------------------------------------------------
> Erik de Castro Lopo
> http://www.mega-nerd.com/
>
> _______________________________________________
> ghc-devs mailing list
> ghc-devs at haskell.org
> http://www.haskell.org/mailman/listinfo/ghc-devs
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20130804/46fd2309/attachment.htm>

Reply | Threaded
Open this post in threaded view
|

libffi

Erik de Castro Lopo-34
Patrick Palka wrote:

> I think the branch check should only be done when performing a pull
> command. That way one can still do ./sync-all fetch, ./sync-all diff, etc
> when on a separate branch.

Yes, that makes sense. Updated patch attached.

Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-sync-all-Bail-out-of-pull-operation-if-not-on-master.patch
Type: text/x-diff
Size: 1038 bytes
Desc: not available
URL: <http://www.haskell.org/pipermail/ghc-devs/attachments/20130804/1f57cde9/attachment.patch>

Reply | Threaded
Open this post in threaded view
|

Multiple repositories, submodules, subtrees and whatnot (Was: Re: libffi)

Niklas Larsson
In reply to this post by Nicolas Trangez
> Given the recent debate about using submodules or subtrees, and the
> occasional issues with sync-all: did anyone ever look into using the
> 'repo' tool? Created by Google for Android development (AFAIK), but also
> used by several other projects. See https://code.google.com/p/git-repo/

Con:
- It doesn't work on Windows at all. Which is a showstopper by itself.
- It doesn't actually solve any of the problems with having tightly
coupled code in different repos. It's just a tool where you can snarf
a boatload of repos at once.
- Everything would have to be restructured into side-by-side gits.

2013/8/3, Nicolas Trangez <nicolas at incubaid.com>:

> Given the recent debate about using submodules or subtrees, and the
> occasional issues with sync-all: did anyone ever look into using the
> 'repo' tool? Created by Google for Android development (AFAIK), but also
> used by several other projects. See https://code.google.com/p/git-repo/
>
> Nicolas
>
> On Sat, 2013-08-03 at 15:41 +0000, Simon Peyton-Jones wrote:
>> |  I think you need to do a 'git pull' first to get the latest version of
>> the
>> |  sync-all script and when you run that you should get a libffi-tarballs
>>
>> I'd done that (repeatedly).  BUT, as has often happened before, I'd
>> forgotten that my tree was on a different branch, and pulling of course
>> has no effect on the branch.   Rats.   Note to self: whenever anything odd
>> happens, check you are on 'master'.
>>
>> Sorry for the noise
>>
>> Simon
>>
>> |  -----Original Message-----
>> |  From: ghc-devs [mailto:ghc-devs-bounces at haskell.org] On Behalf Of Erik
>> de
>> |  Castro Lopo
>> |  Sent: 03 August 2013 08:25
>> |  To: ghc-devs at haskell.org
>> |  Subject: Re: libffi
>> |
>> |  Simon Peyton-Jones wrote:
>> |
>> |  > But ./sync-all get doesn't get a libffi repo.  Nor is it listed in
>> ./packages.
>> |  >
>> |  > Does anyone have any idea where this new libffi repo is, and why it
>> isn't listed?
>> |
>> |  Simon,
>> |
>> |  I think you need to do a 'git pull' first to get the latest version of
>> the
>> |  sync-all script and when you run that you should get a libffi-tarballs
>> |  directory.
>> |
>> |  Cheers,
>> |  Erik
>> |  --
>> |  ----------------------------------------------------------------------
>> |  Erik de Castro Lopo
>> |  http://www.mega-nerd.com/
>> |
>> |  _______________________________________________
>> |  ghc-devs mailing list
>> |  ghc-devs at haskell.org
>> |  http://www.haskell.org/mailman/listinfo/ghc-devs
>>
>>
>>
>> _______________________________________________
>> ghc-devs mailing list
>> ghc-devs at haskell.org
>> http://www.haskell.org/mailman/listinfo/ghc-devs
>
>
>
> _______________________________________________
> ghc-devs mailing list
> ghc-devs at haskell.org
> http://www.haskell.org/mailman/listinfo/ghc-devs
>



Reply | Threaded
Open this post in threaded view
|

libffi

Jan Stolarek
In reply to this post by Erik de Castro Lopo-34
> Better yet, the sync-all script should detect whether it is on master and
> error out (with a decent error message) if not.
Why? I think it is behaving properly - it fetches master branch from the server and merges it into
local master branch, just as expected. I imagine that switching to master branch in all the
subrepos before doing ./sync-all pull will be annoying. Perhaps a warning will be useful here,
but error seems like too much.

Janek





Reply | Threaded
Open this post in threaded view
|

libffi

Erik de Castro Lopo-34
Jan Stolarek wrote:

> > Better yet, the sync-all script should detect whether it is on master and
> > error out (with a decent error message) if not.
> Why? I think it is behaving properly - it fetches master branch from the server and merges it into
> local master branch, just as expected. I imagine that switching to master branch in all the
> subrepos before doing ./sync-all pull will be annoying. Perhaps a warning will be useful here,
> but error seems like too much.

I downgraded it to a warning and commited it.

Cheers,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/