From: "Enrico Scholz" <enrico.scholz@sigma-chemnitz.de>
To: "Ross Burton" <ross@burtonini.com>
Cc: bitbake-devel@lists.openembedded.org
Subject: Re: [bitbake-devel] [PATCH v3 3/4] fetch2/wget: ensure all variables are set when calling urllib
Date: Tue, 17 Aug 2021 20:45:07 +0200 [thread overview]
Message-ID: <ly35r7ykp8.fsf@ensc-virt.intern.sigma-chemnitz.de> (raw)
In-Reply-To: <ly5yw4rmx4.fsf@ensc-virt.intern.sigma-chemnitz.de> (Enrico Scholz via lists openembedded org's message of "Tue, 17 Aug 2021 19:39:03 +0200")
"Enrico Scholz via lists.openembedded.org"
<enrico.scholz=sigma-chemnitz.de@lists.openembedded.org> writes:
> "Ross Burton" <ross@burtonini.com> writes:
>
>> Instead of just exporting the proxy variables when calling into urllib,
>> use bb.utils.environment() to export all of the known variables that are
>> needed for proper connectivity.
>
> This patch breaks fetching from sstate servers[1].
>
> | https_proxy=http://www-cache:3128/
> | no_proxy='.sigma-chemnitz.de,...'
>
> I can reproduce it here e.g. by
>
> | $ bitbake gettext-minimal-native -ccleansstate
> | $ bitbake gettext-minimal-native
> | ERROR: gettext-minimal-native-0.21-r0 do_populate_lic_setscene: No suitable staging package found
> |
> | DEBUG: checkstatus() urlopen failed: <urlopen error Tunnel connection failed: 403 Forbidden>
I guess this happens because in
| + with bb.utils.environment(**newenv):
| + handlers = [...
| + urllib.request.ProxyHandler(),
the urllib.request.ProxyHandler constructor takes 'https_proxy' from
environment (but not 'no_proxy').
Real open happens later (when the 'with bb.utils.environment...' has
been left) and there,
| class ProxyHandler(BaseHandler):
| def proxy_open(self, req, proxy, type):
| ...
| if req.host and proxy_bypass(req.host):
| return None
is called. 'proxy_bypass()' checks 'no_proxy' from env which is not
available anymore.
Normale fetching does not seem to be affected because proxy vars are in
BB_ENV_EXTRAWHITE and exported to jobs. But 'sstate' fetching does not
seem to use this mechanism.
Enrico
next prev parent reply other threads:[~2021-08-17 18:45 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-10 16:55 [PATCH v3 1/4] utils: add environment updating context manager Ross Burton
2021-08-10 16:55 ` [PATCH v3 2/4] fetch2: expose environment variable names that need to be exported Ross Burton
2021-08-10 16:55 ` [PATCH v3 3/4] fetch2/wget: ensure all variables are set when calling urllib Ross Burton
2021-08-17 17:39 ` [bitbake-devel] " Enrico Scholz
2021-08-17 17:44 ` Enrico Scholz
2021-08-17 18:45 ` Enrico Scholz [this message]
2021-08-18 9:28 ` Enrico Scholz
2021-08-18 10:07 ` Ross Burton
2021-08-10 16:55 ` [PATCH v3 4/4] fetch2/wget: fetch securely by default Ross Burton
2021-08-11 13:46 ` [bitbake-devel] " Michael Opdenacker
2021-08-11 15:39 ` Peter Kjellerstedt
2021-08-13 16:41 ` Ross Burton
2021-08-13 22:00 ` Richard Purdie
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=ly35r7ykp8.fsf@ensc-virt.intern.sigma-chemnitz.de \
--to=enrico.scholz@sigma-chemnitz.de \
--cc=bitbake-devel@lists.openembedded.org \
--cc=ross@burtonini.com \
/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.