On Sun, Apr 14, 2019 at 11:42:31PM -0400, kai.k...@windriver.com wrote:
> From: Kai Kang <kai.k...@windriver.com>
> 
> Update SRCREV of cryptodev which only contains one fix for linux 5.0:
>...
> --- a/meta/recipes-kernel/cryptodev/cryptodev.inc
> +++ b/meta/recipes-kernel/cryptodev/cryptodev.inc
> @@ -4,7 +4,7 @@ LICENSE = "GPLv2"
>  LIC_FILES_CHKSUM = "file://COPYING;md5=b234ee4d69f5fce4486a80fdaf4a4263"
>  
>  SRC_URI = "git://github.com/cryptodev-linux/cryptodev-linux"
> -SRCREV = "fd8b15ef1c8398a69a37932ee48c74ab40329a29"
> +SRCREV = "f971e0cd4a0ebe59fb2e8e17240399bf6901b09b"
>  
>  S = "${WORKDIR}/git"

This moves cryptodev from a release to a git snapshot,
but no PV is set to show that.

It is a systematic problem caused by the switch to git hashes for 
releases that random git snapshots look exactly the same as releases
without any clean way to see what a commit actually is.

IMHO this needs support for tags, and then checks like (pseudo-code)
  if(TAG) (commit(TAG) == SRCREV)
  TAG xor (PV contains "+git") 

Short-term for the suggested change in cryptodev the
"which only contains one fix" is correct, and this
is not much different from adding the fix as .patch
which also wouldn't change PV.
So no objection from me to applying the suggested patch as-is.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

-- 
_______________________________________________
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

Reply via email to