[PATCH umr] Skip ahead if PDE entry is actually a PTE entry.

[Christian König]

Close, but not perfect. Reading an address seem to work, but the 
--vm-decode output is a bit confusing:

> [VERBOSE]: PDE0=0x00000003fefff001, VA=0x000000000000, 
> PBA==0x0003fefff000, V=1, S=0, C=0, P=0
> [VERBOSE]:    \-> PDE1=0x00000003feffe001, VA=0x400000000000, 
> PBA==0x0003feffe000, V=1, S=0, C=0, P=0
> [VERBOSE]:       \-> PDE2=0x00000003feffd001, VA=0x000000000000, 
> PBA==0x0003feffd000, V=1, S=0, C=0, P=0
> [VERBOSE]:          \-> PDE3=0x0040000394c004f7, VA=0x000000200000, 
> PBA==0x000394c00000, V=1, S=1, C=1, P=1
> [VERBOSE]:             \-> PTE==0x0040000394c004f7, VA=0x000000018000, 
> PBA==0x000394c00000, V=1, S=1

PDE3 actually doesn't exists and instead it's the PTE pointing to the 
resulting address.

Regards,
Christian.

Am 04.11.2017 um 00:54 schrieb Tom St Denis:
> Signed-off-by: Tom St Denis <tom.stdenis at amd.com>
> ---
>   src/lib/read_vram.c | 6 ++++++
>   1 file changed, 6 insertions(+)
>
> diff --git a/src/lib/read_vram.c b/src/lib/read_vram.c
> index 0df48dadec12..5d9523476e25 100644
> --- a/src/lib/read_vram.c
> +++ b/src/lib/read_vram.c
> @@ -522,6 +522,11 @@ static int umr_access_vram_ai(struct umr_asic *asic, uint32_t vmid,
>   							(int)pde_fields.pte);
>   				memcpy(&pde_array[pde_cnt++], &pde_fields, sizeof pde_fields);
>   
> +				if (pde_fields.pte) {
> +					pte_entry = pde_entry;
> +					goto pde_is_pte;
> +				}
> +
>   				if (!pde_fields.system)
>   					pde_fields.pte_base_addr -= vm_fb_offset;
>   
> @@ -539,6 +544,7 @@ static int umr_access_vram_ai(struct umr_asic *asic, uint32_t vmid,
>   				return -1;
>   
>   			// decode PTE values
> +pde_is_pte:
>   			pte_fields.page_base_addr = pte_entry & 0xFFFFFFFFFF000ULL;
>   			pte_fields.fragment       = (pte_entry >> 7)  & 0x1F;
>   			pte_fields.system         = (pte_entry >> 1) & 1;