On Tue, Jan 19, 2016 at 12:48 AM, Peter Maydell
<peter.mayd...@linaro.org> wrote:
> On 19 January 2016 at 01:33, Alistair Francis
> <alistair.fran...@xilinx.com> wrote:
>> Update the GIC ID registers (registers above 0xfe0) based on the GIC
>> revision instead of using the sames values for all GIC implementations.
>>
>> Signed-off-by: Alistair Francis <alistair.fran...@xilinx.com>
>> Tested-by: Sören Brinkmann <soren.brinkm...@xilinx.com>
>> ---
>>
>> hw/intc/arm_gic.c | 29 ++++++++++++++++++++++++++---
>> 1 file changed, 26 insertions(+), 3 deletions(-)
>>
>> diff --git a/hw/intc/arm_gic.c b/hw/intc/arm_gic.c
>> index 13e297d..f6bfa53 100644
>> --- a/hw/intc/arm_gic.c
>> +++ b/hw/intc/arm_gic.c
>> @@ -31,8 +31,16 @@ do { fprintf(stderr, "arm_gic: " fmt , ## __VA_ARGS__); }
>> while (0)
>> #define DPRINTF(fmt, ...) do {} while(0)
>> #endif
>>
>> -static const uint8_t gic_id[] = {
>> - 0x90, 0x13, 0x04, 0x00, 0x0d, 0xf0, 0x05, 0xb1
>> +static const uint8_t gic_id_11mpcore[] = {
>> + 0x00, 0x00, 0x00, 0x00, 0x90, 0x13, 0x04, 0x00, 0x0d, 0xf0, 0x05, 0xb1
>> +};
>> +
>> +static const uint8_t gic_id_gicv1[] = {
>> + 0x04, 0x00, 0x00, 0x00, 0x90, 0xb3, 0x1b, 0x00, 0x0d, 0xf0, 0x05, 0xb1
>> +};
>> +
>> +static const uint8_t gic_id_gicv2[] = {
>> + 0x04, 0x00, 0x00, 0x00, 0x90, 0xb4, 0x2b, 0x00, 0x0d, 0xf0, 0x05, 0xb1
>> };
>>
>> static inline int gic_get_current_cpu(GICState *s)
>> @@ -689,7 +697,22 @@ static uint32_t gic_dist_readb(void *opaque, hwaddr
>> offset, MemTxAttrs attrs)
>> if (offset & 3) {
>> res = 0;
>> } else {
>> - res = gic_id[(offset - 0xfe0) >> 2];
>> + switch (s->revision) {
>> + case REV_11MPCORE:
>> + res = gic_id_11mpcore[(offset - 0xfe0) >> 2];
>> + break;
>> + case 1:
>> + res = gic_id_gicv1[(offset - 0xfe0) >> 2];
>> + break;
>> + case 2:
>> + res = gic_id_gicv2[(offset - 0xfe0) >> 2];
>> + break;
>> + case REV_NVIC:
>> + /* Shouldn't be able to get here */
>> + abort();
>> + default:
>> + res = 0;
>> + }
>> }
>> }
>> return res;
>
> You've expanded the arrays to include the fd0...fdc values
> (which is right) but the logic also needs to change to
> make offset == 0xfd0..0xfdf go through this code path and
> also to use the new indexing into the array.
I see what you mean, fixing it now.
Thanks,
Alistair
>
> thanks
> -- PMM
>
