Re: [Xen-devel] [v3 for 4.5 2/2] make: Make "src-tarball" target actually make a source tarball

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 16.09.14 at 11:37, <George.Dunlap@xxxxxxxxxxxxx> wrote:
> On Tue, Sep 16, 2014 at 9:55 AM, Jan Beulich <JBeulich@xxxxxxxx> wrote:
>>>>> On 16.09.14 at 10:49, <George.Dunlap@xxxxxxxxxxxxx> wrote:
>>> On Tue, Sep 16, 2014 at 9:11 AM, Jan Beulich <JBeulich@xxxxxxxx> wrote:
>>>>>>> On 15.09.14 at 18:25, <george.dunlap@xxxxxxxxxxxxx> wrote:
>>>>> --- /dev/null
>>>>> +++ b/tools/misc/mktarball
>>>>> @@ -0,0 +1,38 @@
>>>>> +#!/bin/bash
>>>>> +#
>>>>> +# mktarball: Make a release tarball (including xen, qemu, and
>>> qemu-traditional)
>>>>> +#
>>>>> +# Takes 2 arguments, the path to the dist directory and the version
>>>>> +set -ex
>>>>> +
>>>>> +function git_archive_into {
>>>>> +    mkdir "$2"
>>>>> +
>>>>> +    git --git-dir="$1"/.git \
>>>>> +     archive --format=tar HEAD | \
>>>>
>>>> Is HEAD really the right thing here?
>>>
>>> As opposed to...?
>>
>> My git-speak isn't good enough to be sure about suggesting an
>> alternative, but it would seem to me that the respective
>> *_REVISION variables from ./Config.mk should be used here.
>> Iiuc HEAD may refer to a newer commit.
> 
> HEAD should be the current state of the tree, and that's what we want:
> just as the build will take the current state of the git subtrees, we
> want the mktarball script by default to take the current state of
> whatever the git subtrees are.
> 
> But the "make src-tarball[-release]" at the moment will first call
> subtree-force-update on all the trees, so by the time we run the
> script HEAD will be pointing to the _REVISION variables from
> Config.mk.

Ah, okay. As said, my git-speak is pretty limited, so never mind
the comment then.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel