r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-14 Thread Alexander Fyodorov
13.10.2014, 21:50, "Alex Deucher" : > On Fri, Oct 10, 2014 at 12:00 PM, Alex Deucher > wrote: >> ?I think I'd prefer to just switch the test to use gart memory since >> ?this code is shared by different asics thay may not all implement hdp >> ?flush the same way. ?We can just reserve a couple of

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-13 Thread Alex Deucher
On Fri, Oct 10, 2014 at 12:00 PM, Alex Deucher wrote: > On Thu, Oct 9, 2014 at 3:10 PM, Alexander Fyodorov wrote: >> 09.10.2014, 22:32, "Christian K?nig" : >>> Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: 09.10.2014, 21:42, "Christian K?nig" : > For VRAM it is true that we have

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-10 Thread Christian König
Am 10.10.2014 um 18:02 schrieb Alex Deucher: > On Fri, Oct 10, 2014 at 12:00 PM, Alex Deucher > wrote: >> On Thu, Oct 9, 2014 at 3:10 PM, Alexander Fyodorov >> wrote: >>> 09.10.2014, 22:32, "Christian K?nig" : Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: > 09.10.2014, 21:42,

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-10 Thread Alex Deucher
On Fri, Oct 10, 2014 at 12:00 PM, Alex Deucher wrote: > On Thu, Oct 9, 2014 at 3:10 PM, Alexander Fyodorov wrote: >> 09.10.2014, 22:32, "Christian K?nig" : >>> Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: 09.10.2014, 21:42, "Christian K?nig" : > For VRAM it is true that we have

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-10 Thread Alex Deucher
On Thu, Oct 9, 2014 at 3:10 PM, Alexander Fyodorov wrote: > 09.10.2014, 22:32, "Christian K?nig" : >> Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: >>> 09.10.2014, 21:42, "Christian K?nig" : For VRAM it is true that we have a couple of different caches between the CPU and the

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-10 Thread Alexander Fyodorov
09.10.2014, 22:32, "Christian K?nig" : > Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: >> ?09.10.2014, 21:42, "Christian K?nig" : >>> ?For VRAM it is true that we have a couple of different caches between >>> ?the CPU and the actually memory, which need to be flushed explicitly if >>> ?you

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-09 Thread Alexander Fyodorov
09.10.2014, 21:42, "Christian K?nig" : > Hi Alexander, > > in the ring test we write the value 0xDEADBEEF and 0xCAFEDEAD into > registers, not VRAM. > > And the register bar shouldn't be accessed write combined, cause that > could lead to a couple of ordering problems. Why do you think the access

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-09 Thread Christian König
Am 09.10.2014 um 20:15 schrieb Alexander Fyodorov: > 09.10.2014, 21:42, "Christian K?nig" : >> Hi Alexander, >> >> in the ring test we write the value 0xDEADBEEF and 0xCAFEDEAD into >> registers, not VRAM. >> >> And the register bar shouldn't be accessed write combined, cause that >> could lead to

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-09 Thread Christian König
Hi Alexander, in the ring test we write the value 0xDEADBEEF and 0xCAFEDEAD into registers, not VRAM. And the register bar shouldn't be accessed write combined, cause that could lead to a couple of ordering problems. Why do you think the access is done write combined? For VRAM it is true

r600_dma_ring_test() failed - synchronization problem with write-combining memory

2014-10-09 Thread Alexander Fyodorov
Hi David, I'm using 3.10.53-rt56 kernel and encounter a problem in r600_dma_ring_test() when vram memory is mapped as write-combining: no matter how long the polling is done, old value (0xCAFEDEAD) is read. Looking with hardware analyzer at what actually happens in the PCI-E bus, the memory is