On Wed, Oct 19, 2011 at 06:59:54AM -0700, Ronald S. Bultje wrote:
> Hi,
> 
> 2011/10/19 Måns Rullgård <[email protected]>:
> > Janne Grunau <[email protected]> writes:
> >
> >> On Wed, Oct 19, 2011 at 10:55:41AM +0100, Måns Rullgård wrote:
> >>> Kostya Shishkov <[email protected]> writes:
> >>>
> >>> > On Wed, Oct 19, 2011 at 10:35:50AM +0100, Måns Rullgård wrote:
> >>> >> "Ronald S. Bultje" <[email protected]> writes:
> >>> >>
> >>> >> > ---
> >>> >> >  libavcodec/arm/h264dsp_init_arm.c    |    2 +
> >>> >> >  libavcodec/h264.c                    |  126 +++++++++------------
> >>> >> >  libavcodec/h264dsp.c                 |   28 ++----
> >>> >> >  libavcodec/h264dsp.h                 |   10 +-
> >>> >> >  libavcodec/h264dsp_template.c        |   28 ++---
> >>> >> >  libavcodec/ppc/h264_altivec.c        |   44 +++----
> >>> >> >  libavcodec/x86/h264_weight.asm       |  210 
> >>> >> > +++++++++++++---------------------
> >>> >> >  libavcodec/x86/h264_weight_10bit.asm |  145 ++++++++---------------
> >>> >> >  libavcodec/x86/h264dsp_mmx.c         |  175 
> >>> >> > ++++++++++------------------
> >>> >> >  9 files changed, 293 insertions(+), 475 deletions(-)
> >>> >>
> >>> >> Patch does not apply.
> >>> >
> >>> > As discussed on IRC it depends on undisclosed amount of other patches
> >>> > touching H.264.
> >>>
> >>> I think patches sent to the list should apply cleanly.  Review is
> >>> impossible otherwise.
> >>
> >> I agree. If they depend on previous patches still waiting for review,
> >> the entire series should be send.
> >
> > This particular patch shouldn't depend on any others.
> 
> I inverted how it applies, i.e. first "add hacky 422", then "add 422
> inter", then "fix stuff wrong after pushing 422-inter", this is one of
> those.
> 
> I can reorder that, but this patchset has grown too big so it would be
> a lot of effort, for nothing basically, it's one of the typical
> problems of applying stuff so late and having patches linger around.
> Easiest is indeed as Janne did, if I push my stuff to topic branches
> in github and point you guys there, but when I offered that to Kostya,
> he didn't particularly care...
> 
> The same is true for my swscale patches, they are long series and
> everything depends on each other (and my patches currently depend on
> Kieran's patches), so this will be tricky for other people to work
> with. Again, I can push my work to topic branches but I'm not
> particularly keen on reordering patches.

So either roll them as a single patchset or wait till the base is applied,
don't multiply unresolved depndencies.
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to