[PATCH] mirrors.bbclass: Add new KERNELORG_MIRROR = ftp.sunet.se

[PATCH] mirrors.bbclass: Add new KERNELORG_MIRROR = ftp.sunet.se

[Ulf Samuelsson]

www.kernel.org is missing a lot of files.
Add the "ftp.sunet.se" mirror containing many missing files.
The other mirrors in mirrors.bbclass seems to be inactive
so this mirror is added before the inactive mirrors

Signed-off-by: Ulf Samuelsson <ulf@emagii.com>
---
 meta/classes/mirrors.bbclass |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/meta/classes/mirrors.bbclass b/meta/classes/mirrors.bbclass
index 2146087..1124967 100644
--- a/meta/classes/mirrors.bbclass
+++ b/meta/classes/mirrors.bbclass
@@ -23,6 +23,7 @@ ${GNU_MIRROR}	ftp://ftp.cs.ubc.ca/mirror2/gnu \n \
 ${GNU_MIRROR}	ftp://sunsite.ust.hk/pub/gnu \n \
 ${GNU_MIRROR}	ftp://ftp.ayamura.org/pub/gnu \n \
 ${KERNELORG_MIRROR}	http://www.kernel.org/pub \n \
+${KERNELORG_MIRROR}	ftp://ftp.sunet.se/pub/Linux/kernel.org \n \
 ${KERNELORG_MIRROR}	ftp://ftp.us.kernel.org/pub \n \
 ${KERNELORG_MIRROR}	ftp://ftp.uk.kernel.org/pub \n \
 ${KERNELORG_MIRROR}	ftp://ftp.hk.kernel.org/pub \n \
-- 
1.7.5.4

Re: [PATCH] mirrors.bbclass: Add new KERNELORG_MIRROR = ftp.sunet.se

[Koen Kooi]

Op 29 jan. 2012, om 15:26 heeft Ulf Samuelsson het volgende geschreven:

> www.kernel.org is missing a lot of files.
> Add the "ftp.sunet.se" mirror containing many missing files.

Since the files on kernel.org were removed because they might be compromised, why do we add a mirror that has kept the compromised files?

Re: [PATCH] mirrors.bbclass: Add new KERNELORG_MIRROR = ftp.sunet.se

[Richard Purdie]

On Sun, 2012-01-29 at 16:35 +0100, Koen Kooi wrote:
> Op 29 jan. 2012, om 15:26 heeft Ulf Samuelsson het volgende geschreven:
> 
> > www.kernel.org is missing a lot of files.
> > Add the "ftp.sunet.se" mirror containing many missing files.
> 
> Since the files on kernel.org were removed because they might be
> compromised, why do we add a mirror that has kept the compromised
> files?

Its a valid question.

In our favour, we have checksums on all the SRC_URIs these days and I
look suspiciously at anyone changing then other than at version upgrade
time. We therefore have more protection in place than most for this kind
of problem.

Cheers,

Richard