riscv/cpu: add av_const for VLEN helpers

This is read from a read-only constant system register.

libavcodec/riscv/pixblockdsp_init.c

@@ -65,15 +65,18 @@ av_cold void ff_pixblockdsp_init_riscv(PixblockDSPContext *c,
     if ((cpu_flags & AV_CPU_FLAG_RVV_I32) && ff_rv_vlen_least(128)) {
         c->diff_pixels = ff_diff_pixels_unaligned_rvv;
         c->diff_pixels_unaligned = ff_diff_pixels_unaligned_rvv;
-    }
 
-    if ((cpu_flags & AV_CPU_FLAG_RVV_I64) && ff_get_rv_vlenb() >= 16) {
         if (!high_bit_depth) {
-            c->get_pixels = ff_get_pixels_8_rvv;
+            c->get_pixels = ff_get_pixels_unaligned_8_rvv;
             c->get_pixels_unaligned = ff_get_pixels_unaligned_8_rvv;
         }
 
-        c->diff_pixels = ff_diff_pixels_rvv;
+        if (cpu_flags & AV_CPU_FLAG_RVV_I64) {
+            if (!high_bit_depth)
+                c->get_pixels = ff_get_pixels_8_rvv;
+
+            c->diff_pixels = ff_diff_pixels_rvv;
+        }
     }
 #endif
 #endif

libavcodec/riscv/pixblockdsp_rvv.S

@@ -24,7 +24,6 @@ func ff_get_pixels_8_rvv, zve64x
         lpad    0
         vsetivli zero, 8, e8, mf2, ta, ma
         li      t0, 8 * 8
-1:
         vlse64.v v16, (a1), a2
         vsetvli zero, t0, e8, m4, ta, ma
         vwcvtu.x.x.v v8, v16
@@ -32,24 +31,24 @@ func ff_get_pixels_8_rvv, zve64x
         ret
 endfunc
 
-func ff_get_pixels_unaligned_8_rvv, zve64x
+func ff_get_pixels_unaligned_8_rvv, zve32x
         lpad    0
-        andi    t1, a1, 7
-        vsetivli zero, 8, e64, m4, ta, ma
-        li      t0, 8 * 8
-        beqz    t1, 1b
-        andi    a1, a1, -8
-        slli    t2, t1, 3
-        addi    t1, a1, 8
-        sub     t3, t0, t2
-        vlse64.v v16, (a1), a2
-        vlse64.v v24, (t1), a2
-        vsrl.vx v16, v16, t2
-        vsll.vx v24, v24, t3
-        vor.vv  v16, v16, v24
-        vsetvli zero, t0, e8, m4, ta, ma
+        li      t2, 8
+        vsetivli     zero, 8, e8, mf2, ta, ma
+1:
+        add     t1, a1, a2
+        vle8.v  v16, (a1)
+        addi    t0, a0, 2 * 8
+        vle8.v  v17, (t1)
+        addi    t2, t2, -2
         vwcvtu.x.x.v v8, v16
+        vwcvtu.x.x.v v9, v17
         vse16.v v8, (a0)
+        add     a1, t1, a2
+        vse16.v v9, (t0)
+        addi    a0, t0, 2 * 8
+        bnez    t2, 1b
+
         ret
 endfunc
 
@@ -67,17 +66,18 @@ endfunc
 
 func ff_diff_pixels_unaligned_rvv, zve32x
         lpad    0
+        li      t3, 8
         vsetivli    zero, 8, e8, mf2, ta, ma
-        vlsseg8e8.v v16, (a1), a3
-        vlsseg8e8.v v24, (a2), a3
+1:
+        vle8.v v16, (a1)
+        add     a1, a1, a3
+        vle8.v v24, (a2)
+        add     a2, a2, a3
         vwsubu.vv   v8, v16, v24
-        vwsubu.vv   v9, v17, v25
-        vwsubu.vv   v10, v18, v26
-        vwsubu.vv   v11, v19, v27
-        vwsubu.vv   v12, v20, v28
-        vwsubu.vv   v13, v21, v29
-        vwsubu.vv   v14, v22, v30
-        vwsubu.vv   v15, v23, v31
-        vsseg8e16.v v8, (a0)
+        addi    t3, t3, -1
+        vse16.v v8, (a0)
+        addi    a0, a0, 2 * 8
+        bnez    t3, 1b
+
         ret
 endfunc

libavcodec/vulkan_encode.c

@@ -464,6 +464,9 @@ static int vulkan_encode_output(AVCodecContext *avctx,
     VkResult ret;
     FFVulkanEncodePicture *vp = base_pic->priv;
     FFVulkanEncodeContext *ctx = avctx->priv_data;
+    FFHWBaseEncodeContext *base_ctx = &ctx->base;
+    AVPacket *pkt_ptr = pkt;
+
     FFVkBuffer *sd_buf = (FFVkBuffer *)vp->pkt_buf->data;
     uint32_t *query_data;
 
@@ -513,20 +516,56 @@ static int vulkan_encode_output(AVCodecContext *avctx,
         vk->FlushMappedMemoryRanges(ctx->s.hwctx->act_dev, 1, &invalidate_buf);
     }
 
-    pkt->data = sd_buf->mapped_mem;
-    pkt->size = vp->slices_offset + /* base offset */
-                query_data[0]       /* secondary offset */ +
-                query_data[1]       /* size */;
+    if (vp->non_independent_frame) {
+        av_assert0(!ctx->prev_buf_ref);
+        size_t prev_buf_size = vp->slices_offset + query_data[0] + query_data[1];
+        ctx->prev_buf_ref = vp->pkt_buf;
+        ctx->prev_buf_size = prev_buf_size;
+        vp->pkt_buf = NULL;
 
-    /* Move reference */
-    pkt->buf = vp->pkt_buf;
-    vp->pkt_buf = NULL;
+        if (vp->tail_size) {
+            if (base_ctx->tail_pkt->size)
+                return AVERROR_BUG;
+
+            ret = ff_get_encode_buffer(avctx, base_ctx->tail_pkt, vp->tail_size, 0);
+            if (ret < 0)
+                return ret;
+
+            memcpy(base_ctx->tail_pkt->data, vp->tail_data, vp->tail_size);
+            pkt_ptr = base_ctx->tail_pkt;
+        }
+    } else {
+        if (ctx->prev_buf_ref) {
+            FFVkBuffer *prev_sd_buf = (FFVkBuffer *)ctx->prev_buf_ref->data;
+            size_t prev_size = ctx->prev_buf_size;
+            size_t size = (vp->slices_offset + query_data[0] + query_data[1]);
+
+            ret = ff_get_encode_buffer(avctx, pkt, prev_size + size, 0);
+            if (ret < 0)
+                return ret;
+
+            memcpy(pkt->data, prev_sd_buf->mapped_mem, prev_size);
+            memcpy(pkt->data + prev_size, sd_buf->mapped_mem, size);
+
+            av_buffer_unref(&ctx->prev_buf_ref);
+            av_buffer_unref(&vp->pkt_buf);
+        } else {
+            pkt->data = sd_buf->mapped_mem;
+            pkt->size = vp->slices_offset + /* base offset */
+                        query_data[0]       /* secondary offset */ +
+                        query_data[1]       /* size */;
+
+            /* Move reference */
+            pkt->buf = vp->pkt_buf;
+            vp->pkt_buf = NULL;
+        }
+    }
 
     av_log(avctx, AV_LOG_DEBUG, "Frame %"PRId64"/%"PRId64 " encoded\n",
            base_pic->display_order, base_pic->encode_order);
 
     return ff_hw_base_encode_set_output_property(&ctx->base, avctx,
-                                                 base_pic, pkt,
+                                                 base_pic, pkt_ptr,
                                                  ctx->codec->flags & VK_ENC_FLAG_NO_DELAY);
 }
 

libavcodec/vulkan_encode.h

@@ -57,6 +57,10 @@ typedef struct FFVulkanEncodePicture {
     FFVkExecContext       *exec;
     AVBufferRef           *pkt_buf;
     int                    slices_offset;
+
+    int non_independent_frame;
+    char tail_data[16];
+    size_t tail_size;
 } FFVulkanEncodePicture;
 
 /**
@@ -192,6 +196,9 @@ typedef struct FFVulkanEncodeContext {
     FFVkExecPool enc_pool;
 
     FFHWBaseEncodePicture *slots[32];
+
+    AVBufferRef *prev_buf_ref;
+    size_t prev_buf_size;
 } FFVulkanEncodeContext;
 
 #define VULKAN_ENCODE_COMMON_OPTIONS \

libavcodec/vulkan_encode_av1.c

@@ -80,6 +80,8 @@ typedef struct VulkanEncodeAV1Context {
     AV1RawOBU seq_hdr_obu;
     AV1RawOBU meta_cll_obu;
     AV1RawOBU meta_mastering_obu;
+    AV1RawOBU hidden_obu;
+    AV1RawOBU tail_obu;
 
     VkVideoEncodeAV1ProfileInfoKHR profile;
 
@@ -172,7 +174,6 @@ static void set_name_slot(int slot, int *slot_indices, uint32_t allowed_idx, int
     av_assert0(0);
 }
 
-
 static int init_pic_params(AVCodecContext *avctx, FFHWBaseEncodePicture *pic,
                            VkVideoEncodeInfoKHR *encode_info)
 {
@@ -542,6 +543,45 @@ static int init_pic_params(AVCodecContext *avctx, FFHWBaseEncodePicture *pic,
         }
     }
 
+    FFVulkanEncodePicture *vp = pic->priv;
+    vp->tail_size = 0;
+    vp->non_independent_frame = pic->encode_order < pic->display_order;
+    if (vp->non_independent_frame) {
+        AV1RawOBU *obu = &enc->hidden_obu;
+        AV1RawFrameHeader *fh = &obu->obu.frame_header;
+
+        /** hidden frame header */
+        memset(obu, 0, sizeof(*obu));
+        obu->header.obu_type = AV1_OBU_FRAME_HEADER;
+        obu->header.obu_has_size_field = 1;
+
+        fh->frame_type = AV1_FRAME_INTER;
+        fh->refresh_frame_flags = 1 << ap->slot;
+        fh->frame_width_minus_1   = base_ctx->surface_width - 1;
+        fh->frame_height_minus_1  = base_ctx->surface_height - 1;
+        fh->render_width_minus_1  = fh->frame_width_minus_1;
+        fh->render_height_minus_1 = fh->frame_height_minus_1;
+
+        memcpy(fh->loop_filter_ref_deltas, default_loop_filter_ref_deltas,
+               AV1_TOTAL_REFS_PER_FRAME * sizeof(int8_t));
+
+        obu = &enc->tail_obu;
+        fh = &obu->obu.frame_header;
+
+        /** tail frame header */
+        memset(obu, 0, sizeof(*obu));
+        obu->header.obu_type = AV1_OBU_FRAME_HEADER;
+        obu->header.obu_has_size_field = 1;
+
+        fh->show_existing_frame   = 1;
+        fh->frame_to_show_map_idx = ap->slot != 0;
+        fh->frame_type            = AV1_FRAME_INTER;
+        fh->frame_width_minus_1   = base_ctx->surface_width - 1;
+        fh->frame_height_minus_1  = base_ctx->surface_height - 1;
+        fh->render_width_minus_1  = fh->frame_width_minus_1;
+        fh->render_height_minus_1 = fh->frame_height_minus_1;
+    }
+
     return 0;
 }
 
@@ -1079,8 +1119,31 @@ static int write_extra_headers(AVCodecContext *avctx,
     int err;
     VulkanEncodeAV1Context *enc = avctx->priv_data;
     VulkanEncodeAV1Picture  *ap = base_pic->codec_priv;
+    FFVulkanEncodePicture *vp = base_pic->priv;
     CodedBitstreamFragment *obu = &enc->current_access_unit;
 
+    if (vp->non_independent_frame) {
+        err = vulkan_encode_av1_add_obu(avctx, obu, AV1_OBU_FRAME_HEADER, &enc->hidden_obu);
+        if (err < 0)
+            goto fail;
+
+        // Only for tracking ref frame in context, not to be output
+        err = ff_cbs_write_fragment_data(enc->cbs, obu);
+        if (err < 0)
+            goto fail;
+
+        ff_cbs_fragment_reset(obu);
+        ((CodedBitstreamAV1Context *)enc->cbs->priv_data)->seen_frame_header = 0;
+
+        err = vulkan_encode_av1_add_obu(avctx, obu, AV1_OBU_FRAME_HEADER, &enc->tail_obu);
+        if (err < 0)
+            goto fail;
+
+        err = vulkan_encode_av1_write_obu(avctx, vp->tail_data, &vp->tail_size, obu);
+        if (err < 0)
+            goto fail;
+    }
+
     if (ap->units_needed & UNIT_MASTERING_DISPLAY) {
         err = vulkan_encode_av1_add_obu(avctx, obu,
                                         AV1_OBU_METADATA,

libavutil/riscv/cpu.h

@@ -45,7 +45,7 @@ static inline av_const bool ff_rv_zbb_support(void)
  * Returns the vector size in bytes (always a power of two and at least 4).
  * This is undefined behaviour if vectors are not implemented.
  */
-static inline size_t ff_get_rv_vlenb(void)
+static inline av_const size_t ff_get_rv_vlenb(void)
 {
     size_t vlenb;
 
@@ -61,7 +61,7 @@ static inline size_t ff_get_rv_vlenb(void)
  * Checks that the vector bit-size is at least the given value.
  * This is potentially undefined behaviour if vectors are not implemented.
  */
-static inline bool ff_rv_vlen_least(unsigned int bits)
+static inline av_const bool ff_rv_vlen_least(unsigned int bits)
 {
 #ifdef __riscv_v_min_vlen
     if (bits <= __riscv_v_min_vlen)