<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On 17 September 2015 at 19:52, Rob Clark <span dir="ltr"><<a href="mailto:robdclark@gmail.com" target="_blank">robdclark@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span>On Wed, Sep 16, 2015 at 5:48 AM, Xinliang Liu <<a href="mailto:xinliang.liu@linaro.org" target="_blank">xinliang.liu@linaro.org</a>> wrote:<br> ><br> ><br> > On 16 September 2015 at 02:25, Rob Herring <<a href="mailto:robh@kernel.org" target="_blank">robh@kernel.org</a>> wrote:<br> > Hi Rob, thanks a lot for reply:-)<br> ><br> >> On 09/15/2015 04:37 AM, Xinwei Kong wrote:<br> >> > If you config DRM_HISI_FBDEV optional, this patch will only support<br> >> > fbdev<br> >> > mode while also supporting double buffer.<br> >><br> >> This is a lot of duplicated code from CMA fbdev. Is double buffering the<br> >> only reason why CMA fbdev can't be used or are there some other<br> >> constraints?<br> ><br> > Yes, double buffering is the main reason we rewrite our own fbdev.<br> > CMA fbdev only create one buffer. But we need at least double buffer for<br> > running Android with fbdev.<br> ><br> >> Double buffering in fbdev has always been a hack, so I'm<br> >> guessing that is not a feature that should be added here.<br> >><br> > If so, I think it is hard to be accepted for my cma fbdev patch to support<br> > multi buffer.<br> > This early week, I have sent a cma fbdev patch for supporting this. The<br> > subject is<br> > "[PATCH] drm/cma-helper: Add multi buffer support for cma fbdev".<br> > We do have a strong will to support this feature. I described the reason in<br> > the patch. Please take a look for me. Thank you very much.<br> <br> </span>fwiw, drm_gralloc has support for kms. </blockquote><div>We do have a plan to replace fbdev with kms. But it take time.</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Currently it is expected to be<br> paired w/ a mesa gpu driver, which might not work for everyone, but I<br> suppose the display part of it could be extracted out for a<br> <a href="http://gralloc.kms.so" rel="noreferrer" target="_blank">gralloc.kms.so</a> for pure sw rendering.. that might be a better<br> approach.<br></blockquote><div> </div><div>Yes, I also think kms should be handle in hwc rather than in gralloc.</div><div>IMO, gralloc should just handle dma buffer alloc and free.</div><div>We have a reference hwc implementation. which will handle kms inside it.</div><div><div style="font-size:12.8000001907349px">wiki: <a href="https://wiki.linaro.org/BenjaminGaignard/HWComposer_DRM?highlight=%28hwcomposer%2" target="_blank">https://wiki.linaro.org/BenjaminGaignard/HWComposer_DRM?highlight=%28hwcomposer%2</a></div><div style="font-size:12.8000001907349px">source code: <span style="color:rgb(51,51,51);font-family:Ubuntu,sans-serif;font-size:13px;line-height:20.7999992370605px">git://<a href="http://git.linaro.org/people/benjamin.gaignard/hwcomposer.git" target="_blank">git.linaro.org/people/benjamin.gaignard/hwcomposer.git</a></span></div></div><div style="font-size:12.8000001907349px"><br></div><div style="font-size:12.8000001907349px"><span style="font-size:12.8000001907349px">Cheers</span>,</div><div style="font-size:12.8000001907349px">-Xinliang</div><div style="font-size:12.8000001907349px"><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"> <br> <a href="http://git.android-x86.org/?p=platform/hardware/drm_gralloc.git" rel="noreferrer" target="_blank">http://git.android-x86.org/?p=platform/hardware/drm_gralloc.git</a><br> <br> BR,<br> -R<br> <span><br> > -Xinliang<br> ><br> >> Rob<br> >><br> >> > Signed-off-by: Xinliang Liu <<a href="mailto:xinliang.liu@linaro.org" target="_blank">xinliang.liu@linaro.org</a>><br> >> > Signed-off-by: Xinwei Kong <<a href="mailto:kong.kongxinwei@hisilicon.com" target="_blank">kong.kongxinwei@hisilicon.com</a>><br> >> > Signed-off-by: Andy Green <<a href="mailto:andy.green@linaro.org" target="_blank">andy.green@linaro.org</a>><br> >> > Signed-off-by: Jiwen Qi <<a href="mailto:qijiwen@hisilicon.com" target="_blank">qijiwen@hisilicon.com</a>><br> >> > Signed-off-by: Yu Gong <<a href="mailto:gongyu@hisilicon.com" target="_blank">gongyu@hisilicon.com</a>><br> >> > ---<br> >> > drivers/gpu/drm/hisilicon/Kconfig | 13 +<br> >> > drivers/gpu/drm/hisilicon/Makefile | 3 +-<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_connector.c | 4 +<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_drv.c | 9 +<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_dsi.c | 15 +<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_fb.h | 5 +<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_fbdev.c | 395<br> >> > +++++++++++++++++++++++++<br> >> > drivers/gpu/drm/hisilicon/hisi_drm_fbdev.h | 24 ++<br> >> > 8 files changed, 467 insertions(+), 1 deletion(-)<br> >> > create mode 100644 drivers/gpu/drm/hisilicon/hisi_drm_fbdev.c<br> >> > create mode 100644 drivers/gpu/drm/hisilicon/hisi_drm_fbdev.h<br> >><br> ><br> ><br> </span>> _______________________________________________<br> > dri-devel mailing list<br> > <a href="mailto:dri-devel@lists.freedesktop.org" target="_blank">dri-devel@lists.freedesktop.org</a><br> > <a href="http://lists.freedesktop.org/mailman/listinfo/dri-devel" rel="noreferrer" target="_blank">http://lists.freedesktop.org/mailman/listinfo/dri-devel</a><br> ><br> </blockquote></div><br></div></div>