On 27.05.2015 14:34, Daniel Vetter wrote: > On Tue, May 26, 2015 at 05:21:56PM -0700, Rodrigo Vivi wrote: >> On Thu, May 21, 2015 at 5:04 AM, Animesh Manna <animesh.manna@xxxxxxxxx> wrote: >>> Naming convention of csr firmware will be - >>> <platform>_dmc_<api-version>_<minor-version>.bin >>> >>> Accordingly updated the same in code. >>> >>> Signed-off-by: Animesh Manna <animesh.manna@xxxxxxxxx> >>> --- >>> drivers/gpu/drm/i915/intel_csr.c | 2 +- >>> 1 file changed, 1 insertion(+), 1 deletion(-) >>> >>> diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c >>> index fec2bc5..9bd05bf 100644 >>> --- a/drivers/gpu/drm/i915/intel_csr.c >>> +++ b/drivers/gpu/drm/i915/intel_csr.c >>> @@ -34,7 +34,7 @@ >>> * low-power state and comes back to normal. >>> */ >>> >>> -#define I915_CSR_SKL "i915/skl_dmc_ver4.bin" >>> +#define I915_CSR_SKL "i915/skl_dmc_1_4.bin" >> >> I believe in this case the install instructions should point to copy >> instead of linking it, but not hardcode the release version. >> >> So shouldn't it be i915/skl_dmc_ver1.bin then? >> >> or i915/skl_dmc_ver1_4.bin in worst case... > > While at it, can we perhaps _not_ hardcode the minor revision? Hardcoding > the minor revision torpedoes the entire concept of even having a minor > revision. If the firmware team can't do proper abi versioning, then imo we > should just put one number for each and keep it at that. > -Daniel Can we please settle on something? I'd rather have the csr fw in linux-firmware quite soon, so that the driver wouldn't give an ugly assert when csr doesn't work. -- t _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
