On Wed, 08 Sep 2021, Dave Airlie <airlied@xxxxxxxxx> wrote:
> From: Dave Airlie <airlied@xxxxxxxxxx>
>
> This reworks the uncore function vtable so that it's constant.
There's a bug in there, see comment inline, with that fixed,
Reviewed-by: Jani Nikula <jani.nikula@xxxxxxxxx>
>
> Signed-off-by: Dave Airlie <airlied@xxxxxxxxxx>
> ---
> drivers/gpu/drm/i915/intel_uncore.c | 139 +++++++++++++++++-----------
> drivers/gpu/drm/i915/intel_uncore.h | 8 +-
> 2 files changed, 89 insertions(+), 58 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
> index d0bbfc320604..0bc6e16fc4e3 100644
> --- a/drivers/gpu/drm/i915/intel_uncore.c
> +++ b/drivers/gpu/drm/i915/intel_uncore.c
> @@ -1756,32 +1756,24 @@ __vgpu_write(8)
> __vgpu_write(16)
> __vgpu_write(32)
>
> -#define ASSIGN_RAW_WRITE_MMIO_VFUNCS(uncore, x) \
> -do { \
> - (uncore)->funcs.mmio_writeb = x##_write8; \
> - (uncore)->funcs.mmio_writew = x##_write16; \
> - (uncore)->funcs.mmio_writel = x##_write32; \
> -} while (0)
> -
> -#define ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, x) \
> -do { \
> - (uncore)->funcs.mmio_readb = x##_read8; \
> - (uncore)->funcs.mmio_readw = x##_read16; \
> - (uncore)->funcs.mmio_readl = x##_read32; \
> - (uncore)->funcs.mmio_readq = x##_read64; \
> -} while (0)
> -
> -#define ASSIGN_WRITE_MMIO_VFUNCS(uncore, x) \
> -do { \
> - ASSIGN_RAW_WRITE_MMIO_VFUNCS((uncore), x); \
> - (uncore)->funcs.write_fw_domains = x##_reg_write_fw_domains; \
> -} while (0)
> -
> -#define ASSIGN_READ_MMIO_VFUNCS(uncore, x) \
> -do { \
> - ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, x); \
> - (uncore)->funcs.read_fw_domains = x##_reg_read_fw_domains; \
> -} while (0)
> +#define MMIO_RAW_WRITE_VFUNCS(x) \
> + .mmio_writeb = x##_write8, \
> + .mmio_writew = x##_write16, \
> + .mmio_writel = x##_write32
> +
> +#define MMIO_RAW_READ_VFUNCS(x) \
> + .mmio_readb = x##_read8, \
> + .mmio_readw = x##_read16, \
> + .mmio_readl = x##_read32, \
> + .mmio_readq = x##_read64
> +
> +#define MMIO_WRITE_FW_VFUNCS(x) \
> + MMIO_RAW_WRITE_VFUNCS(x), \
> + .write_fw_domains = x##_reg_write_fw_domains
> +
> +#define MMIO_READ_FW_VFUNCS(x) \
> + MMIO_RAW_READ_VFUNCS(x), \
> + .read_fw_domains = x##_reg_read_fw_domains
>
> static int __fw_domain_init(struct intel_uncore *uncore,
> enum forcewake_domain_id domain_id,
> @@ -2086,22 +2078,70 @@ void intel_uncore_init_early(struct intel_uncore *uncore,
> uncore->debug = &i915->mmio_debug;
> }
>
> +static const struct intel_uncore_funcs vgpu_funcs = {
> + MMIO_RAW_WRITE_VFUNCS(vgpu),
> + MMIO_RAW_READ_VFUNCS(vgpu),
> +};
> +
> +static const struct intel_uncore_funcs gen5_funcs = {
> + MMIO_RAW_WRITE_VFUNCS(gen5),
> + MMIO_RAW_READ_VFUNCS(gen5),
> +};
> +
> +static const struct intel_uncore_funcs gen2_funcs = {
> + MMIO_RAW_WRITE_VFUNCS(gen2),
> + MMIO_RAW_READ_VFUNCS(gen2),
> +};
> +
> +
> static void uncore_raw_init(struct intel_uncore *uncore)
> {
> GEM_BUG_ON(intel_uncore_has_forcewake(uncore));
>
> if (intel_vgpu_active(uncore->i915)) {
> - ASSIGN_RAW_WRITE_MMIO_VFUNCS(uncore, vgpu);
> - ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, vgpu);
> + uncore->funcs = &vgpu_funcs;
> } else if (GRAPHICS_VER(uncore->i915) == 5) {
> - ASSIGN_RAW_WRITE_MMIO_VFUNCS(uncore, gen5);
> - ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, gen5);
> + uncore->funcs = &gen5_funcs;
> } else {
> - ASSIGN_RAW_WRITE_MMIO_VFUNCS(uncore, gen2);
> - ASSIGN_RAW_READ_MMIO_VFUNCS(uncore, gen2);
> + uncore->funcs = &gen2_funcs;
> }
> }
>
> +static const struct intel_uncore_funcs xehp_funcs = {
> + MMIO_WRITE_FW_VFUNCS(xehp_fwtable),
> + MMIO_READ_FW_VFUNCS(gen11_fwtable)
> +};
> +
> +static const struct intel_uncore_funcs gen12_funcs = {
> + MMIO_WRITE_FW_VFUNCS(gen12_fwtable),
> + MMIO_READ_FW_VFUNCS(gen12_fwtable)
> +};
> +
> +static const struct intel_uncore_funcs gen11_funcs = {
> + MMIO_WRITE_FW_VFUNCS(gen11_fwtable),
> + MMIO_READ_FW_VFUNCS(gen11_fwtable)
> +};
> +
> +static const struct intel_uncore_funcs gen9_funcs = {
> + MMIO_WRITE_FW_VFUNCS(fwtable),
> + MMIO_READ_FW_VFUNCS(fwtable)
> +};
> +
> +static const struct intel_uncore_funcs gen8_funcs = {
> + MMIO_WRITE_FW_VFUNCS(gen8),
> + MMIO_READ_FW_VFUNCS(gen6)
> +};
> +
> +static const struct intel_uncore_funcs vlv_funcs = {
> + MMIO_WRITE_FW_VFUNCS(gen8),
Should be gen6.
> + MMIO_READ_FW_VFUNCS(fwtable)
> +};
> +
> +static const struct intel_uncore_funcs gen6_funcs = {
> + MMIO_WRITE_FW_VFUNCS(gen6),
> + MMIO_READ_FW_VFUNCS(gen6)
> +};
> +
> static int uncore_forcewake_init(struct intel_uncore *uncore)
> {
> struct drm_i915_private *i915 = uncore->i915;
> @@ -2116,38 +2156,29 @@ static int uncore_forcewake_init(struct intel_uncore *uncore)
>
> if (GRAPHICS_VER_FULL(i915) >= IP_VER(12, 55)) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __dg2_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, xehp_fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen11_fwtable);
> + uncore->funcs = &xehp_funcs;
> } else if (GRAPHICS_VER_FULL(i915) >= IP_VER(12, 50)) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __xehp_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, xehp_fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen11_fwtable);
> + uncore->funcs = &xehp_funcs;
> } else if (GRAPHICS_VER(i915) >= 12) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __gen12_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen12_fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen12_fwtable);
> + uncore->funcs = &gen12_funcs;
> } else if (GRAPHICS_VER(i915) == 11) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __gen11_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen11_fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen11_fwtable);
> + uncore->funcs = &gen11_funcs;
> } else if (IS_GRAPHICS_VER(i915, 9, 10)) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __gen9_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, fwtable);
> + uncore->funcs = &gen9_funcs;
> } else if (IS_CHERRYVIEW(i915)) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __chv_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, fwtable);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, fwtable);
> + uncore->funcs = &gen9_funcs;
Seems funny to use gen9 functions for chv.
> } else if (GRAPHICS_VER(i915) == 8) {
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen8);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen6);
> + uncore->funcs = &gen8_funcs;
> } else if (IS_VALLEYVIEW(i915)) {
> ASSIGN_FW_DOMAINS_TABLE(uncore, __vlv_fw_ranges);
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen6);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, fwtable);
> + uncore->funcs = &vlv_funcs;
> } else if (IS_GRAPHICS_VER(i915, 6, 7)) {
> - ASSIGN_WRITE_MMIO_VFUNCS(uncore, gen6);
> - ASSIGN_READ_MMIO_VFUNCS(uncore, gen6);
> + uncore->funcs = &gen6_funcs;
> }
>
> uncore->pmic_bus_access_nb.notifier_call = i915_pmic_bus_access_notifier;
> @@ -2190,8 +2221,8 @@ int intel_uncore_init_mmio(struct intel_uncore *uncore)
>
> /* make sure fw funcs are set if and only if we have fw*/
> GEM_BUG_ON(intel_uncore_has_forcewake(uncore) != !!uncore->fw_get_funcs);
> - GEM_BUG_ON(intel_uncore_has_forcewake(uncore) != !!uncore->funcs.read_fw_domains);
> - GEM_BUG_ON(intel_uncore_has_forcewake(uncore) != !!uncore->funcs.write_fw_domains);
> + GEM_BUG_ON(intel_uncore_has_forcewake(uncore) != !!uncore->funcs->read_fw_domains);
> + GEM_BUG_ON(intel_uncore_has_forcewake(uncore) != !!uncore->funcs->write_fw_domains);
Gah, I hate the proliferation of GEM_BUG_ON() outside of strictly gem
code. If there's value in it beyond gem, it should not be named
GEM. Otherwise, it should stay within GEM code.
>
> if (HAS_FPGA_DBG_UNCLAIMED(i915))
> uncore->flags |= UNCORE_HAS_FPGA_DBG_UNCLAIMED;
> @@ -2530,10 +2561,10 @@ intel_uncore_forcewake_for_reg(struct intel_uncore *uncore,
> return 0;
>
> if (op & FW_REG_READ)
> - fw_domains = uncore->funcs.read_fw_domains(uncore, reg);
> + fw_domains = uncore->funcs->read_fw_domains(uncore, reg);
>
> if (op & FW_REG_WRITE)
> - fw_domains |= uncore->funcs.write_fw_domains(uncore, reg);
> + fw_domains |= uncore->funcs->write_fw_domains(uncore, reg);
>
> drm_WARN_ON(&uncore->i915->drm, fw_domains & ~uncore->fw_domains);
>
> diff --git a/drivers/gpu/drm/i915/intel_uncore.h b/drivers/gpu/drm/i915/intel_uncore.h
> index 1a7448f5f16f..52d4baf07656 100644
> --- a/drivers/gpu/drm/i915/intel_uncore.h
> +++ b/drivers/gpu/drm/i915/intel_uncore.h
> @@ -138,7 +138,7 @@ struct intel_uncore {
>
> struct notifier_block pmic_bus_access_nb;
> const struct intel_uncore_fw_get *fw_get_funcs;
> - struct intel_uncore_funcs funcs;
> + const struct intel_uncore_funcs *funcs;
>
> unsigned int fifo_count;
>
> @@ -312,14 +312,14 @@ __raw_write(64, q)
> static inline u##x__ intel_uncore_##name__(struct intel_uncore *uncore, \
> i915_reg_t reg) \
> { \
> - return uncore->funcs.mmio_read##s__(uncore, reg, (trace__)); \
> + return uncore->funcs->mmio_read##s__(uncore, reg, (trace__)); \
> }
>
> #define __uncore_write(name__, x__, s__, trace__) \
> static inline void intel_uncore_##name__(struct intel_uncore *uncore, \
> i915_reg_t reg, u##x__ val) \
> { \
> - uncore->funcs.mmio_write##s__(uncore, reg, val, (trace__)); \
> + uncore->funcs->mmio_write##s__(uncore, reg, val, (trace__)); \
> }
>
> __uncore_read(read8, 8, b, true)
> @@ -338,7 +338,7 @@ __uncore_write(write_notrace, 32, l, false)
> * an arbitrary delay between them. This can cause the hardware to
> * act upon the intermediate value, possibly leading to corruption and
> * machine death. For this reason we do not support intel_uncore_write64,
> - * or uncore->funcs.mmio_writeq.
> + * or uncore->funcs->mmio_writeq.
> *
> * When reading a 64-bit value as two 32-bit values, the delay may cause
> * the two reads to mismatch, e.g. a timestamp overflowing. Also note that
--
Jani Nikula, Intel Open Source Graphics Center
