checksums...

checksums...

[Koen Kooi]

Hi,

Could people adding new recipe and people git-am'ing patches adding new 
recipes please add the corresponding checksums to checksums.ini?

It's not hard:

1) bitbake <recipe>
2) watch it fail
3) cd $TMPDIR ; cat /OE/org.openembedded.dev/conf/checksums.ini >> 
checksums.ini ; python 
/OE/org.openembedded.dev/contrib/source-checker/oe-checksums-sorter.py 
checksums.ini > /OE/org.openembedded.dev/conf/checksums.ini
4) git commit --amend /OE/org.openembedded.dev/conf/checksums.ini

regards,

Koen

Re: checksums...

[Philip Balister]

[-- Attachment #1: Type: text/plain, Size: 834 bytes --]

Koen Kooi wrote:
> Hi,
> 
> Could people adding new recipe and people git-am'ing patches adding new 
> recipes please add the corresponding checksums to checksums.ini?
> 
> It's not hard:
> 
> 1) bitbake <recipe>
> 2) watch it fail
> 3) cd $TMPDIR ; cat /OE/org.openembedded.dev/conf/checksums.ini >> 
> checksums.ini ; python 
> /OE/org.openembedded.dev/contrib/source-checker/oe-checksums-sorter.py 
> checksums.ini > /OE/org.openembedded.dev/conf/checksums.ini
> 4) git commit --amend /OE/org.openembedded.dev/conf/checksums.ini

I screwed this up once (and committed a recipe without the checksum) 
when I left ALLOW_INSECURE_DOWNLOADS set in my local.conf. Now I have it 
commented out and am really careful only to use it only when hacking.

I also need to look up what "git commit --amend" does.

Philip

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/x-pkcs7-signature, Size: 3303 bytes --]

[RFC] Add option for strict checksumming, was: Re: checksums...

[Koen Kooi]

On 21-01-09 20:15, Philip Balister wrote:
> Koen Kooi wrote:
>> Hi,
>>
>> Could people adding new recipe and people git-am'ing patches adding
>> new recipes please add the corresponding checksums to checksums.ini?
>>
>> It's not hard:
>>
>> 1) bitbake <recipe>
>> 2) watch it fail
>> 3) cd $TMPDIR ; cat /OE/org.openembedded.dev/conf/checksums.ini >>
>> checksums.ini ; python
>> /OE/org.openembedded.dev/contrib/source-checker/oe-checksums-sorter.py
>> checksums.ini > /OE/org.openembedded.dev/conf/checksums.ini
>> 4) git commit --amend /OE/org.openembedded.dev/conf/checksums.ini
>
> I screwed this up once (and committed a recipe without the checksum)
> when I left ALLOW_INSECURE_DOWNLOADS set in my local.conf. Now I have it
> commented out and am really careful only to use it only when hacking.

What about this:

 From a48ee59ac6ea569a5f88066d86de4b66d8f606c3 Mon Sep 17 00:00:00 2001
From: Koen Kooi <koen@openembedded.org>
Date: Wed, 21 Jan 2009 20:29:53 +0100
Subject: [PATCH] base.bbclass: only error out on missing checksums when 
OE_STRICT_CHECKSUMS is set
  * we want to error out on non-matching sums, missing ones are less bad

---
  classes/base.bbclass |    8 +++++++-
  1 files changed, 7 insertions(+), 1 deletions(-)

diff --git a/classes/base.bbclass b/classes/base.bbclass
index fca0edd..411adc9 100644
--- a/classes/base.bbclass
+++ b/classes/base.bbclass
@@ -71,7 +71,13 @@ def base_chk_file(parser, pn, pv, src_uri, localpath, 
data):

          file.write("[%s]\nmd5=%s\nsha256=%s\n\n" % (src_uri, md5data, 
shadata))
          file.close()
-        return False
+        if not bb.data.getVar("OE_STRICT_CHECKSUMS",data, True):
+            bb.note("This package has no entry in checksums.ini, please 
add one")
+            bb.note("\n[%s]\nmd5=%s\nsha256=%s" % (src_uri, md5data, 
shadata))
+            return True
+        else:
+            bb.note("Missing checksum")
+            return False

      if not md5 == md5data:
          bb.note("The MD5Sums did not match. Wanted: '%s' and Got: 
'%s'" % (md5,md5data))
-- 
1.5.6.3

Non-mangled version at 
http://dominion.thruhere.net/koen/OE/0001-base.bbclass-only-error-out-on-missing-checksums-wh.patch

regards,

Koen

Re: [RFC] Add option for strict checksumming, was: Re: checksums...

[Tom Rini]

On Wed, Jan 21, 2009 at 08:34:51PM +0100, Koen Kooi wrote:
> On 21-01-09 20:15, Philip Balister wrote:
>> Koen Kooi wrote:
>>> Hi,
>>>
>>> Could people adding new recipe and people git-am'ing patches adding
>>> new recipes please add the corresponding checksums to checksums.ini?
>>>
>>> It's not hard:
>>>
>>> 1) bitbake <recipe>
>>> 2) watch it fail
>>> 3) cd $TMPDIR ; cat /OE/org.openembedded.dev/conf/checksums.ini >>
>>> checksums.ini ; python
>>> /OE/org.openembedded.dev/contrib/source-checker/oe-checksums-sorter.py
>>> checksums.ini > /OE/org.openembedded.dev/conf/checksums.ini
>>> 4) git commit --amend /OE/org.openembedded.dev/conf/checksums.ini
>>
>> I screwed this up once (and committed a recipe without the checksum)
>> when I left ALLOW_INSECURE_DOWNLOADS set in my local.conf. Now I have it
>> commented out and am really careful only to use it only when hacking.
>
> What about this:
>
> From a48ee59ac6ea569a5f88066d86de4b66d8f606c3 Mon Sep 17 00:00:00 2001
> From: Koen Kooi <koen@openembedded.org>
> Date: Wed, 21 Jan 2009 20:29:53 +0100
> Subject: [PATCH] base.bbclass: only error out on missing checksums when  
> OE_STRICT_CHECKSUMS is set
>  * we want to error out on non-matching sums, missing ones are less bad

This will bring us back to where we started.  I think we just need to
live with the pain a bit more and get people used to "Oops, no checksum
in your patch" ...

And btw, git commit --amend will let you change things in the last
commit you made.  ONLY do this if you haven't also done a push.

-- 
Tom Rini