On 08.09.2014 14:50, Axel Davy wrote: > Hi, > > When reading si_dma.c code, it looks like the requested width of the > copy is ignored except for PIPE_BUFFER. > Perhaps that explains the bugs observed ? >
It isn't ignored. Partial DMA copies (i.e. operations that do not copy whole lines) are simply not supported right now, and will fall back to resource_copy_region. In fact, it's even stricter: the picthes of source and destination have to match. Grigori > Axel Davy > > On 06/09/2014 19:14, Andy Furniss wrote : >> Marek Olšák wrote: >>> From: Marek Olšák <[email protected]> >>> >>> --- >>> >>> Turn this on, run piglit, and pray for mercy. >>> It might be interesting to see if it makes 3D apps any faster. Or >>> piglit. >> >> Well it's not piglit and I haven't benchmarked anything yet, but I get >> a couple of faults at the start of Unigine Valley. >> >> [20635.429686] radeon 0000:01:00.0: GPU fault detected: 146 0x07bd3d14 >> [20635.429690] radeon 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR >> 0x0000683D >> [20635.429691] radeon 0000:01:00.0: >> VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x1D03D014 >> [20635.429693] VM fault (0x04, vmid 14) at page 26685, write from DMA1 >> (61) >> [20635.429708] radeon 0000:01:00.0: GPU fault detected: 146 0x07bd3d14 >> [20635.429709] radeon 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR >> 0x00000000 >> [20635.429710] radeon 0000:01:00.0: >> VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x1C03D00C >> [20635.429711] VM fault (0x0c, vmid 14) at page 0, read from DMA1 (61) >> >> _______________________________________________ >> mesa-dev mailing list >> [email protected] >> http://lists.freedesktop.org/mailman/listinfo/mesa-dev > > _______________________________________________ > mesa-dev mailing list > [email protected] > http://lists.freedesktop.org/mailman/listinfo/mesa-dev
signature.asc
Description: OpenPGP digital signature
_______________________________________________ mesa-dev mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/mesa-dev
