Also if I skip this IOCTL in the code then memtester works normaly and say that memory is normal.
пятница, 3 апреля 2015 г., 19:20:38 UTC+3 пользователь Andrew Kosteltsev написал: > > > Hi Siarhei > > I have built u-boot02015.01 and latest stage/sunxi-3.4 kernel (four memory > chips 512Mx8 on the board) and trying to use lima-memspeed utility. > > $ ./lima-memspeed gpu_copy > > GPU jobs works properly but PP jobs crashes the system (I have inserted > printf() functions to your code and ...). > > I have found the point where the system reboots. It happens when we call > LIMA_M400_PP_START_JOB_R3P0.. > > G2P and FB allocation is follow: > > SYS: 0x43000000 - 0x4300FFFF ( 64 KB) > G2D: 0x50000000 - 0x57FFFFFF (128 MB) > LCD: 0x58000000 - 0x5FFFFFFF (128 MB) > > dcdc3 vol 1300000 uV > > system crashes exactly during the IOCTL: > > limadriver/limare/lib/jobs.c~410: > > int limare_m400_pp_job_start_r3p0( > .... > .... > > ret = ioctl( state-fd, LIMA_M400_PP_START_JOB_R3P0, &job); > > > probably this error related to translation physical addresses in the > kernel mali driver when we access to pixel processors. > > Please advise how to fix this error or make some workaround. > > Thanks in advance. > Andrey K. > > > -- You received this message because you are subscribed to the Google Groups "linux-sunxi" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
