One question on wayland desing on buffer
Hi experts: wayland designs a light-weight server, usually client manages buffer by themselves. It gives client more flexibility, migration rendering/composition between different server, subsurface to keep layout while keeping view hiarchy. however, there is something not convenient for wayland server. take android surfaceflinger as counterpart, surfaceflinger manages the buffer/layer of surface, then any client (owns the BufferQueue client proxy) can update the surface directly. that is, mediaserver in android can enqueueBuffer to update the surface directly. in wayland world, a mediaserver has to pass the buffer (handle) to app/UI process, then the surface is updated in app process. Is my above understanding correct? ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
回复:Re: bo flags isn't passed through wayland-drm protocl, how weston respect these flags?
some buffer attributes may influence how weston use it, for example: tiling mode, Swizzling, compression.weston may depend on these attribute to decide to assign the buffer to a hw plane or composite it as texture. - 原始邮件 - 发件人:Pekka Paalanen 收件人:"HalleyZhao" 抄送人:"wayland-devel" 主题:Re: bo flags isn't passed through wayland-drm protocl, how weston respect these flags? 日期:2019年08月06日 15点57分 On Tue, 06 Aug 2019 09:03:51 +0800 "HalleyZhao" wrote: > Hi experts:when we create buffer object at wayland client side, there are > some usage flags, for example gbm_bo_flags.but when we pass these buffer fd > to weston through wayland-drm protocl, these flags are ignored.then, how > weston respect these flags when using the buffer object?thanks. > > > type="new_id" interface="wl_buffer"/> > name="format" type="uint"/> name="stride0" type="int"/> name="stride1" type="int"/> name="stride2" type="int"/> > > Hi, why should the compositor care about usage flags at all? What are you thinking about here? The buffer has already been allocated by the client. The primary purpose of the usage flags is to guide the allocation, so the compositor has no use for them. Compositor's usage of the buffer on the other hand is implied by the Wayland protocol. If you attach the wl_buffer to a wl_surface, then the buffer will be used as a texture, possibly as a KMS FB, and maybe more if it turns out suitable (Weston implements only the texture and KMS FB cases). If you use the wl_buffer in some other Wayland interface, then that other interface implies how it will be used. Scott is also right. gbm_bo_flags predate DRM format modifiers. gbm_bo_flags guide the GBM implementation in choosing the modifier, and the modifier will be sent to the Wayland compositor to describe the data layout of the buffer in zwp_linux_dmabuf extension. As you can see, gbm_bo_create_with_modifiers() does not take gbm_bo_flags at all. Thanks, pq ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
bo flags isn't passed through wayland-drm protocl, how weston respect these flags?
Hi experts:when we create buffer object at wayland client side, there are some usage flags, for example gbm_bo_flags.but when we pass these buffer fd to weston through wayland-drm protocl, these flags are ignored.then, how weston respect these flags when using the buffer object?thanks. ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
回复:Re: Re: Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1?
thanks,I'm clear now - 原始邮件 - 发件人:Daniel Stone 收件人:halleyz...@sina.com 抄送人:daniels , wayland-devel 主题:Re: Re: Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1? 日期:2019年07月25日 17点22分 On Thu, 25 Jul 2019 at 10:17, HalleyZhao wrote: > looking for an example to use in EGL Weston itself provides an example of using dma-fences with EGL - both extracting them from EGL to refer to previous rendering, and asking EGL to wait on them. If you're looking for an example of how you would modify your EGL implementation to use the explicit-sync extension, then there is no useful example since it depends heavily on how your EGL implementation is structured internally. Cheers, Daniel ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
回复:Re: Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1?
looking for an example to use in EGLthanks - 原始邮件 - 发件人:Daniel Stone 收件人:halleyz...@sina.com 抄送人:daniels , wayland-devel 主题:Re: Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1? 日期:2019年07月25日 13点45分 On Thu, 25 Jul 2019 at 04:34, HalleyZhao wrote: > in Mesa, I could find some implement for vulkan, but not for EGL. > > maybe, desktop developer care more on mesa than mobile developers, and most > desktop driver does implicitly sync instead of depending on > explicit-synchronization-unstable-v1. 'Mobile' GL drivers in Mesa also do implicit synchronisation currently. Are you looking for Mesa usage because you need to fix a specific issue on some platform? Or are you just looking for an example to use? I don't really understand how to help here. Cheers, Daniel ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
回复:Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1?
thanks Simon. in Mesa, I could find some implement for vulkan, but not for EGL. maybe, desktop developer care more on mesa than mobile developers, and most desktop driver does implicitly sync instead of depending on explicit-synchronization-unstable-v1. - 原始邮件 - 发件人:Simon Ser 收件人:"halleyz...@sina.com" 抄送人:daniels , wayland-devel 主题:Re:_回复:does_mesa_support_linux-explicit-synchronization-unstable-v1? 日期:2019年07月24日 14点02分 On Wednesday, July 24, 2019 3:07 AM, HalleyZhao wrote: > I noticed that weston merged some patches to support explicited GPU > synchronization: > https://bugs.freedesktop.org/show_bug.cgi?id=98766 > > when I tried to search corresponding feature in mesa, nothing is found. > is this feature test-able with recent weston/mesa? > thanks. Hi, I think you need to manually write the glue code between the EGL extension and the Wayland protocol. See the example client in clients/simple-dmabuf-egl.c. -- Simon Ser https://emersion.fr ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
回复:does mesa support linux-explicit-synchronization-unstable-v1?
Hi Daniel:could you tell me the status of linux-explicit-synchronization-unstable-v1 in mesa? - 原始邮件 - 发件人:"HalleyZhao" 收件人:"wayland-devel" 主题:does mesa support linux-explicit-synchronization-unstable-v1? 日期:2019年07月23日 14点05分 Hi experts:I noticed that weston merged some patches to support explicited GPU synchronization: https://bugs.freedesktop.org/show_bug.cgi?id=98766 when I tried to search corresponding feature in mesa, nothing is found.is this feature test-able with recent weston/mesa?thanks. ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel
does mesa support linux-explicit-synchronization-unstable-v1?
Hi experts:I noticed that weston merged some patches to support explicited GPU synchronization: https://bugs.freedesktop.org/show_bug.cgi?id=98766 when I tried to search corresponding feature in mesa, nothing is found.is this feature test-able with recent weston/mesa?thanks. ___ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel