Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-23 Thread Daniel Vetter
On Mon, Feb 22, 2021 at 06:30:35PM +0100, Noralf Trønnes wrote: > > > Den 22.02.2021 08.54, skrev Thomas Zimmermann: > > Hi > > > > Am 20.02.21 um 14:02 schrieb Noralf Trønnes: > >> > >> > >> Den 19.02.2021 14.40, skrev Thomas Zimmermann: > >>> Fixes a regression for udl and probably other USB-b

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-22 Thread Noralf Trønnes
Den 22.02.2021 08.54, skrev Thomas Zimmermann: > Hi > > Am 20.02.21 um 14:02 schrieb Noralf Trønnes: >> >> >> Den 19.02.2021 14.40, skrev Thomas Zimmermann: >>> Fixes a regression for udl and probably other USB-based drivers where >>> joining and mirroring displays fails. >>> >>> Joining display

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-21 Thread Thomas Zimmermann
Hi Am 20.02.21 um 14:02 schrieb Noralf Trønnes: Den 19.02.2021 14.40, skrev Thomas Zimmermann: Fixes a regression for udl and probably other USB-based drivers where joining and mirroring displays fails. Joining displays requires importing a dma_buf from another DRM driver. These devices don'

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-21 Thread Thomas Zimmermann
Hi Am 19.02.21 um 16:53 schrieb Alan Stern: On Fri, Feb 19, 2021 at 02:45:54PM +0100, Christian König wrote: Well as far as I can see this is a relative clear NAK. When a device can't do DMA and has no DMA mask then why it is requesting an sg-table in the first place? This may not be importa

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-21 Thread Christian König
Am 20.02.21 um 14:02 schrieb Noralf Trønnes: Den 19.02.2021 14.40, skrev Thomas Zimmermann: Fixes a regression for udl and probably other USB-based drivers where joining and mirroring displays fails. Joining displays requires importing a dma_buf from another DRM driver. These devices don't s

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-20 Thread Noralf Trønnes
Den 19.02.2021 14.40, skrev Thomas Zimmermann: > Fixes a regression for udl and probably other USB-based drivers where > joining and mirroring displays fails. > > Joining displays requires importing a dma_buf from another DRM driver. > These devices don't support DMA and therefore have no DMA m

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Alan Stern
On Fri, Feb 19, 2021 at 04:56:16PM +0100, Christian König wrote: > > > Am 19.02.21 um 16:53 schrieb Alan Stern: > > On Fri, Feb 19, 2021 at 02:45:54PM +0100, Christian König wrote: > > > Well as far as I can see this is a relative clear NAK. > > > > > > When a device can't do DMA and has no DMA

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Alan Stern
On Fri, Feb 19, 2021 at 02:45:54PM +0100, Christian König wrote: > Well as far as I can see this is a relative clear NAK. > > When a device can't do DMA and has no DMA mask then why it is requesting an > sg-table in the first place? This may not be important for your discussion, but I'd like to g

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Christian König
Am 19.02.21 um 16:53 schrieb Alan Stern: On Fri, Feb 19, 2021 at 02:45:54PM +0100, Christian König wrote: Well as far as I can see this is a relative clear NAK. When a device can't do DMA and has no DMA mask then why it is requesting an sg-table in the first place? This may not be important

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Daniel Vetter
On Fri, Feb 19, 2021 at 3:35 PM Thomas Zimmermann wrote: > > Hi > > Am 19.02.21 um 14:45 schrieb Christian König: > > Well as far as I can see this is a relative clear NAK. > > > > When a device can't do DMA and has no DMA mask then why it is requesting > > an sg-table in the first place? > > > >

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Christian König
Am 19.02.21 um 15:34 schrieb Thomas Zimmermann: Hi Am 19.02.21 um 14:45 schrieb Christian König: Well as far as I can see this is a relative clear NAK. When a device can't do DMA and has no DMA mask then why it is requesting an sg-table in the first place? The problem is rather in drm_gem

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Thomas Zimmermann
Hi Am 19.02.21 um 14:45 schrieb Christian König: Well as far as I can see this is a relative clear NAK. When a device can't do DMA and has no DMA mask then why it is requesting an sg-table in the first place? The problem is rather in drm_gem_prime_import_dev() which always tries to create a

Re: [PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Christian König
Well as far as I can see this is a relative clear NAK. When a device can't do DMA and has no DMA mask then why it is requesting an sg-table in the first place? The problem is rather in drm_gem_prime_import_dev() which always tries to create an sg-table to get a GEM object, even when the devic

[PATCH] drm/prime: Only call dma_map_sgtable() for devices with DMA support

2021-02-19 Thread Thomas Zimmermann
Fixes a regression for udl and probably other USB-based drivers where joining and mirroring displays fails. Joining displays requires importing a dma_buf from another DRM driver. These devices don't support DMA and therefore have no DMA mask. Trying to map imported buffers from a DMA-able memory z