On Thu, Jul 11, 2019 at 03:10:04PM +0100, Chris Wilson wrote:
> With the fancy second slice of Haswell, we lost support for Ivybridge
> with its paltry single slice.
>
> Fixes: 8ddcfd6882a9 ("intel_l3_parity: slice support")
> Testcase: igt/tools_test/sysfs_l3_parity
> Signed-off-by: Chris Wilson <[email protected]>
Looks sane.
I wonder if anyone ever used this feature...
> ---
> tools/intel_l3_parity.c | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/tools/intel_l3_parity.c b/tools/intel_l3_parity.c
> index d8c997af9..dc3d08048 100644
> --- a/tools/intel_l3_parity.c
> +++ b/tools/intel_l3_parity.c
> @@ -195,6 +195,9 @@ int main(int argc, char *argv[])
>
> for_each_slice(i) {
> fd[i] = openat(dir, path[i], O_RDWR);
> + if (fd[i] < 0)
> + continue;
> +
> if (read(fd[i], l3logs[i], NUM_REGS * sizeof(uint32_t)) < 0) {
> perror(path[i]);
> exit(77);
> @@ -333,6 +336,9 @@ int main(int argc, char *argv[])
>
> /* Per slice operations */
> for_each_slice(i) {
> + if (fd[i] < 0)
> + continue;
> +
> switch (action) {
> case 'l':
> dumpit(i);
> @@ -374,6 +380,9 @@ int main(int argc, char *argv[])
> exit(EXIT_SUCCESS);
>
> for_each_slice(i) {
> + if (fd[i] < 0)
> + continue;
> +
> ret = write(fd[i], l3logs[i], NUM_REGS * sizeof(uint32_t));
> if (ret == -1) {
> perror("Writing sysfs");
> --
> 2.22.0
>
> _______________________________________________
> Intel-gfx mailing list
> [email protected]
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
--
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
