Re: [PATCH 5.11 regression fix] drm/ttm: fix combining __GFP_HIGHMEM and __GFP_DMA32 flag for DMA32 pools

2021-01-13 Thread Hans de Goede
Hi, On 1/13/21 9:25 AM, Christian König wrote: > Hi Hans, > > Am 12.01.21 um 19:32 schrieb Hans de Goede: >> GFP_TRANSHUGE_LIGHT includes __GFP_HIGHMEM and combining >> __GFP_HIGHMEM with __GFP_DMA32 is not allowed. >> >> So we must not set add GFP_TRANSHUGE_LIGHT to the gfp_flags when >> allocat

Re: [PATCH 5.11 regression fix] drm/ttm: fix combining __GFP_HIGHMEM and __GFP_DMA32 flag for DMA32 pools

2021-01-13 Thread Christian König
Hi Hans, Am 12.01.21 um 19:32 schrieb Hans de Goede: GFP_TRANSHUGE_LIGHT includes __GFP_HIGHMEM and combining __GFP_HIGHMEM with __GFP_DMA32 is not allowed. So we must not set add GFP_TRANSHUGE_LIGHT to the gfp_flags when allocating pages from a dma32 pool. This won't work since we still need

Re: [PATCH 5.11 regression fix] drm/ttm: fix combining __GFP_HIGHMEM and __GFP_DMA32 flag for DMA32 pools

2021-01-12 Thread Huang Rui
On Wed, Jan 13, 2021 at 02:32:49AM +0800, Hans de Goede wrote: > GFP_TRANSHUGE_LIGHT includes __GFP_HIGHMEM and combining > __GFP_HIGHMEM with __GFP_DMA32 is not allowed. > > So we must not set add GFP_TRANSHUGE_LIGHT to the gfp_flags when > allocating pages from a dma32 pool. > > This fixes the

[PATCH 5.11 regression fix] drm/ttm: fix combining __GFP_HIGHMEM and __GFP_DMA32 flag for DMA32 pools

2021-01-12 Thread Hans de Goede
GFP_TRANSHUGE_LIGHT includes __GFP_HIGHMEM and combining __GFP_HIGHMEM with __GFP_DMA32 is not allowed. So we must not set add GFP_TRANSHUGE_LIGHT to the gfp_flags when allocating pages from a dma32 pool. This fixes the following oops when using a driver which uses DMA32 pools such as the vboxvid