From: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
To: Alejandro del Castillo <alejandro.delcastillo@ni.com>
Cc: openembedded-core@lists.openembedded.org
Subject: Re: [oe][PATCH 2/2] package_manager: support for signed IPK package feeds
Date: Wed, 18 Nov 2015 12:01:18 +0200 [thread overview]
Message-ID: <20151118120118.48fcec99@adi-pc-linux> (raw)
In-Reply-To: <564B928A.1090005@ni.com>
On Tue, 17 Nov 2015 14:48:10 -0600
Alejandro del Castillo <alejandro.delcastillo@ni.com> wrote:
>
>
> On 11/17/2015 09:26 AM, Ioan-Adrian Ratiu wrote:
> > Create gpg signed package feeds if configured. Very similar to
> > how rpm does it. Most of the config variables are shared with
> > the rpm backend (like PACKAGE_FEED_GPG_NAME), with the exception
> > of PACKAGE_FEED_GPG_PUBKEY which is not needed in this case.
> >
> > Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
> > ---
> > meta/lib/oe/package_manager.py | 23 ++++++++++++++++++++++-
> > 1 file changed, 22 insertions(+), 1 deletion(-)
> >
> > diff --git a/meta/lib/oe/package_manager.py
> > b/meta/lib/oe/package_manager.py index 964fddc..8528c9b 100644
> > --- a/meta/lib/oe/package_manager.py
> > +++ b/meta/lib/oe/package_manager.py
> > @@ -174,10 +174,25 @@ class OpkgIndexer(Indexer):
> >
> > opkg_index_cmd = bb.utils.which(os.getenv('PATH'),
> > "opkg-make-index")
> > + gpg_cmd = ''
> > +
> > + # all these variables are needed to succesfully sign the
> > index, otherwise skip signing
> > + if self.d.getVar('PACKAGE_FEED_SIGN', True) == '1' and \
> > + self.d.getVar('PACKAGE_FEED_GPG_NAME', True) and \
> > + self.d.getVar('PACKAGE_FEED_GPG_PASSPHRASE_FILE', True):
> > + pkgfeed_gpg_name =
> > self.d.getVar('PACKAGE_FEED_GPG_NAME', True)
> > + pkgfeed_gpg_pass =
> > self.d.getVar('PACKAGE_FEED_GPG_PASSPHRASE_FILE', True)
> > + gpg_bin = self.d.getVar('GPG_BIN', True) or
> > bb.utils.which(os.getenv('PATH'), "gpg") +
> > + gpg_cmd = "%s --no-use-agent --batch --yes -ab -u
> > %s --passphrase-file '%s'" % \
> > + (gpg_bin, pkgfeed_gpg_name,
> > pkgfeed_gpg_pass) +
> > +
>
> I think you can combine this block with the "if gpg_cmd:" one below
> (move this logic to the bottom and combine it with the content of the
> if gpg_cmd block)
>
> if not os.path.exists(os.path.join(self.deploy_dir,
> "Packages")):
> > open(os.path.join(self.deploy_dir, "Packages"),
> > "w").close()
> > index_cmds = []
> > + index_sign_files = []
> > for arch_var in arch_vars:
> > archs = self.d.getVar(arch_var, True)
> > if archs is None:
> > @@ -196,6 +211,8 @@ class OpkgIndexer(Indexer):
> > index_cmds.append('%s -r %s -p %s -m %s' %
> > (opkg_index_cmd, pkgs_file,
> > pkgs_file, pkgs_dir))
> > + index_sign_files.append(pkgs_file)
> > +
> > if len(index_cmds) == 0:
> > bb.note("There are no packages in %s!" %
> > self.deploy_dir) return
> > @@ -206,7 +223,11 @@ class OpkgIndexer(Indexer):
> > if self.d.getVar('PACKAGE_FEED_SIGN', True) == '1':
> > raise NotImplementedError('Package feed signing not
> > implementd for ipk')
>
> Not need anymore
>
> > -
> > + if gpg_cmd:
> > + for f in index_sign_files:
> > + result = oe.utils.multiprocess_exec([gpg_cmd + ' '
> > + f], create_index)
> > + if result:
> > + bb.fatal('%s' % ('\n'.join(result)))
> >
> > class DpkgIndexer(Indexer):
> > def _create_configs(self):
> >
>
Writing this way was a choice to avoid code duplication and looping the
arches a second time. If I were to put all logic under the same
condition at the bottom as you suggest, I will have to duplicate all
code that recreates the pkgs_file values needed for signing.
Instead, IMO a better idea is to combine the first block (if
self.d.getVar('PACKAGE_FEED_SIGN'...) with the last one (if gpg_cmd) as
you suggest but keep the "index_sign_files.append(pkgs_file)" line to
avoid looping all the arches a second time and duplicate all that code.
I'll resubmit a v2 to exemplify what I'm saying here.
Question: Is there a better way to get the index file names without
looping through the arches? If yes, then all this code can be put under
a single if branch as you suggest and I agree this is the best case
scenario.
next prev parent reply other threads:[~2015-11-18 10:01 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-17 15:26 [oe][PATCH 1/2] package_ipk: support signing of ipk packages Ioan-Adrian Ratiu
2015-11-17 15:26 ` [oe][PATCH 2/2] package_manager: support for signed IPK package feeds Ioan-Adrian Ratiu
2015-11-17 20:48 ` Alejandro del Castillo
2015-11-18 10:01 ` Ioan-Adrian Ratiu [this message]
2015-11-18 10:25 ` [oe][PATCH v2 1/2] package_ipk: support signing of ipk packages Ioan-Adrian Ratiu
2015-11-18 10:25 ` [oe][PATCH v2 2/2] package_manager: support for signed IPK package feeds Ioan-Adrian Ratiu
2015-11-18 16:00 ` Alejandro del Castillo
2015-11-19 8:35 ` Ioan-Adrian Ratiu
2015-11-19 15:29 ` Alejandro del Castillo
2015-11-19 15:41 ` [oe][PATCH v3 1/2] package_ipk: support signing of ipk packages Ioan-Adrian Ratiu
2015-11-19 15:41 ` [oe][PATCH v3 2/2] package_manager: support for signed IPK package feeds Ioan-Adrian Ratiu
2015-11-19 15:58 ` Alejandro del Castillo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20151118120118.48fcec99@adi-pc-linux \
--to=adrian.ratiu@ni.com \
--cc=alejandro.delcastillo@ni.com \
--cc=openembedded-core@lists.openembedded.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.