On 9/20/2019 5:48 PM, Jani Nikula wrote:
On Fri, 20 Sep 2019, Animesh Manna <animesh.manna@xxxxxxxxx> wrote:
DSB can program large set of data through indexed register write
(opcode 0x9) in one shot. DSB feature can be used for bulk register
programming e.g. gamma lut programming, HDR meta data programming.
v1: initial version.
v2: simplified code by using ALIGN(). (Chris)
v3: ascii table added as code comment. (Shashank)
v4: cosmetic changes done. (Shashank)
v5: reset ins_start_offset. (Jani)
v6: update ins_start_offset in inel_dsb_reg_write.
Cc: Shashank Sharma <shashank.sharma@xxxxxxxxx>
Cc: Imre Deak <imre.deak@xxxxxxxxx>
Cc: Jani Nikula <jani.nikula@xxxxxxxxx>
Cc: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>
Reviewed-by: Shashank Sharma <shashank.sharma@xxxxxxxxx>
Signed-off-by: Animesh Manna <animesh.manna@xxxxxxxxx>
---
drivers/gpu/drm/i915/display/intel_dsb.c | 68 ++++++++++++++++++++++++
drivers/gpu/drm/i915/display/intel_dsb.h | 9 ++++
2 files changed, 77 insertions(+)
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c b/drivers/gpu/drm/i915/display/intel_dsb.c
index f94cd6dc98b6..faa853b08458 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.c
+++ b/drivers/gpu/drm/i915/display/intel_dsb.c
@@ -12,8 +12,10 @@
/* DSB opcodes. */
#define DSB_OPCODE_SHIFT 24
#define DSB_OPCODE_MMIO_WRITE 0x1
+#define DSB_OPCODE_INDEXED_WRITE 0x9
#define DSB_BYTE_EN 0xF
#define DSB_BYTE_EN_SHIFT 20
+#define DSB_REG_VALUE_MASK 0xfffff
struct intel_dsb *
intel_dsb_get(struct intel_crtc *crtc)
@@ -83,9 +85,74 @@ void intel_dsb_put(struct intel_dsb *dsb)
mutex_unlock(&i915->drm.struct_mutex);
dsb->cmd_buf = NULL;
dsb->free_pos = 0;
+ dsb->ins_start_offset = 0;
}
}
+void intel_dsb_indexed_reg_write(struct intel_dsb *dsb, i915_reg_t reg,
+ u32 val)
+{
+ struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
+ struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+ u32 *buf = dsb->cmd_buf;
+ u32 reg_val;
+
+ if (!buf) {
+ I915_WRITE(reg, val);
+ return;
+ }
+
+ if (WARN_ON(dsb->free_pos >= DSB_BUF_SIZE)) {
+ DRM_DEBUG_KMS("DSB buffer overflow\n");
+ return;
+ }
+
+ /*
+ * For example the buffer will look like below for 3 dwords for auto
+ * increment register:
+ * +--------------------------------------------------------+
+ * | size = 3 | offset &| value1 | value2 | value3 | zero |
+ * | | opcode | | | | |
+ * +--------------------------------------------------------+
+ * + + + + + + +
+ * 0 4 8 12 16 20 24
+ * Byte
+ *
+ * As every instruction is 8 byte aligned the index of dsb instruction
+ * will start always from even number while dealing with u32 array. If
+ * we are writing odd no of dwords, Zeros will be added in the end for
+ * padding.
+ */
+ reg_val = buf[dsb->ins_start_offset + 1] & DSB_REG_VALUE_MASK;
+ if (reg_val != i915_mmio_reg_offset(reg)) {
+ /* Every instruction should be 8 byte aligned. */
+ dsb->free_pos = ALIGN(dsb->free_pos, 2);
+
+ dsb->ins_start_offset = dsb->free_pos;
+
+ /* Update the size. */
+ buf[dsb->free_pos++] = 1;
+
+ /* Update the opcode and reg. */
+ buf[dsb->free_pos++] = (DSB_OPCODE_INDEXED_WRITE <<
+ DSB_OPCODE_SHIFT) |
+ i915_mmio_reg_offset(reg);
+
+ /* Update the value. */
+ buf[dsb->free_pos++] = val;
+ } else {
+ /* Update the new value. */
+ buf[dsb->free_pos++] = val;
+
+ /* Update the size. */
+ buf[dsb->ins_start_offset]++;
+ }
+
+ /* if number of data words is odd, then the last dword should be 0.*/
+ if (dsb->free_pos & 0x1)
+ buf[dsb->free_pos] = 0;
+}
+
void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val)
{
struct intel_crtc *crtc = container_of(dsb, typeof(*crtc), dsb);
@@ -102,6 +169,7 @@ void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val)
return;
}
+ dsb->ins_start_offset = dsb->free_pos;
Okay, I'm being a pedant, but that's kind of part of the job
description, I'm afraid.
What if:
intel_dsb_get()
intel_dsb_reg_write(dsb, FOO, 0);
intel_dsb_indexed_reg_write(dsb, FOO, 0);
intel_dsb_commit()
intel_dsb_put()
Hi Jani,
I am trying to think a scenario where may write the same register which
is having auto-increment capability using both intel_dsb_reg_write and
intel_dsb_indexed_reg_write.
To set the auto increment mode we may need to write a different register
to control auto-increment mode.
If there is any practical scenario, do you want to me to add now?
Currently checking register value only while creating buffer for
auto-increment register. If we want to add the above then we might need
like below because we are introducing a variability factor regarding
dsb-api to write the same register in consequent call.
enum dsb_write_type {
NORMAL,
INDEX
};
struct last_write {
u32 reg_val;
enum dsb_write_type type;
}
then, last write can be updated in intel_dsb_reg_write() as NORMAL write
and later will check in intel_dsb_indexed_reg_write() to identify above
mentioned case.
Please let me know your suggestion, will do accordingly.
Regards,
Animesh
BR,
Jani.
buf[dsb->free_pos++] = val;
buf[dsb->free_pos++] = (DSB_OPCODE_MMIO_WRITE << DSB_OPCODE_SHIFT) |
(DSB_BYTE_EN << DSB_BYTE_EN_SHIFT) |
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.h b/drivers/gpu/drm/i915/display/intel_dsb.h
index 0686d67b34d5..2ae22f7309a7 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.h
+++ b/drivers/gpu/drm/i915/display/intel_dsb.h
@@ -30,11 +30,20 @@ struct intel_dsb {
* and help in calculating tail of command buffer.
*/
int free_pos;
+
+ /*
+ * ins_start_offset will help to store start address of the dsb
+ * instuction and help in identifying the batch of auto-increment
+ * register.
+ */
+ u32 ins_start_offset;
};
struct intel_dsb *
intel_dsb_get(struct intel_crtc *crtc);
void intel_dsb_put(struct intel_dsb *dsb);
void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val);
+void intel_dsb_indexed_reg_write(struct intel_dsb *dsb, i915_reg_t reg,
+ u32 val);
#endif
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
