--- Begin Message ---
Subject: |
[PATCH] gnu: libva: Update to 2.13.0 |
Date: |
Mon, 1 Nov 2021 16:03:48 +0000 |
* gnu/packages/video.scm (libva): Update it.
---
gnu/packages/video.scm | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/gnu/packages/video.scm b/gnu/packages/video.scm
index d4d9eed68d..7f86dcc10c 100644
--- a/gnu/packages/video.scm
+++ b/gnu/packages/video.scm
@@ -1425,14 +1425,14 @@ (define-public libmatroska
(define-public libva
(package
(name "libva")
- (version "2.10.0")
+ (version "2.13.0")
(source
(origin
(method url-fetch)
(uri (string-append "https://github.com/intel/libva/releases/download/"
version "/libva-" version ".tar.bz2"))
(sha256
- (base32 "0dh2zjn6wi74ga75r6pbrrj8hjm213zyxvn9bv78z0fra1dy70gs"))))
+ (base32 "0q6l193x9whd80sjd5mx8cb7c0fcljb19nhfpla5h49nkzrq7lzs"))))
(build-system gnu-build-system)
(native-inputs
`(("pkg-config" ,pkg-config)))
--
2.33.1
--- End Message ---
--- Begin Message ---
Subject: |
Re: bug#51554: [PATCH] gnu: libva: Update to 2.13.0 |
Date: |
Thu, 25 Nov 2021 01:49:46 -0500 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/27.2 (gnu/linux) |
Hello,
Josselin Poiret <dev@jpoiret.xyz> writes:
> Ludovic Courtès <ludo@gnu.org> writes:
>> For the record, this entails lots of rebuilds, contrary to what one
>> might think:
>>
>> --8<---------------cut here---------------start------------->8---
>> $ ./pre-inst-env guix refresh -l -e '(force (@@ (gnu packages gl)
>> libva-without-mesa))'
>> Building the following 1478 packages would ensure 2711 dependent packages
>> are rebuilt: […]
>> --8<---------------cut here---------------end--------------->8---
>>
>> I suppose this could go to ‘core-updates’.
>
> My bad, I completely missed this. I initially wanted to upgrade libva
> because the older version doesn't work on core-updates-frozen with
> Wayland compositors, so I think it should ideally belong there if
> possible to avoid having broken hardware acceleration for a while when
> that branch is merged. If it does indeed lead to too many rebuilds,
> then maybe it could go onto core-updates-frozen-batched-changes, but
> then that'd delay its merging again. I'm not too sure between those,
> but I think it should be one of them.
I had totally missed that for the batched branch.
Too bad. I know Ludovic had concerns about a ~3K rebuilds; that's a lot
but currently only x86_64 and i686 are in a state of building the world
it seems, so that's about 6K packages.
I've bunched this with glib-networking, another similar sized rebuild,
manually built most of the dependents on berlin for x86_64, and pushed
as 0a787e67ecd0fab42c7ddfec639b53d58c15af6e.
Closing.
Maxim
--- End Message ---