Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
 help / color / mirror / Atom feed
From: remi@remlab.net
To: ffmpeg-devel@ffmpeg.org
Subject: [FFmpeg-devel] [PATCH 05/26] lavu/cpu: CPU flags for the RISC-V Vector extension
Date: Tue, 20 Sep 2022 17:39:52 +0300
Message-ID: <20220920144013.4959-5-remi@remlab.net> (raw)
In-Reply-To: <5602865.DvuYhMxLoT@basile.remlab.net>

From: Rémi Denis-Courmont <remi@remlab.net>

RVV defines a total of 12 different extensions, including:

- 5 different instruction subsets:
  - Zve32x: 8-, 16- and 32-bit integers,
  - Zve32f: Zve32x plus single precision floats,
  - Zve64x: Zve32x plus 64-bit integers,
  - Zve64f: Zve32f plus Zve64x,
  - Zve64d: Zve64f plus double precision floats.

- 6 different vector lengths:
  - Zvl32b (embedded only),
  - Zvl64b (embedded only),
  - Zvl128b,
  - Zvl256b,
  - Zvl512b,
  - Zvl1024b,

- and the V extension proper: equivalent to Zve64f and Zvl128b.

In total, there are 6 different possible sets of supported instructions
(including the empty set), but for convenience we allocate one bit for
each type sets: up-to-32-bit ints (ZVE32X), floats (ZV32F),
64-bit ints (ZV64X) and doubles (ZVE64D).

Whence the vector size is needed, it can be retrieved by reading the
unprivileged read-only vlenb CSR. This should probably be a separate
helper macro if needed at a later point.
---
 libavutil/cpu.c           |  4 ++++
 libavutil/cpu.h           |  4 ++++
 libavutil/riscv/cpu.c     | 46 ++++++++++++++++++++++++++++++++++++++-
 tests/checkasm/checkasm.c | 10 ++++++---
 4 files changed, 60 insertions(+), 4 deletions(-)

diff --git a/libavutil/cpu.c b/libavutil/cpu.c
index 78e92a1bf6..58ae4858b4 100644
--- a/libavutil/cpu.c
+++ b/libavutil/cpu.c
@@ -187,6 +187,10 @@ int av_parse_cpu_caps(unsigned *flags, const char *s)
         { "rvi",      NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVI      },    .unit = "flags" },
         { "rvf",      NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVF      },    .unit = "flags" },
         { "rvd",      NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVD      },    .unit = "flags" },
+        { "rvve32",   NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RV_ZVE32X},    .unit = "flags" },
+        { "rvvf",     NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RV_ZVE32F},    .unit = "flags" },
+        { "rvve64",   NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RV_ZVE64X},    .unit = "flags" },
+        { "rvv",      NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RV_ZVE64D},    .unit = "flags" },
 #endif
         { NULL },
     };
diff --git a/libavutil/cpu.h b/libavutil/cpu.h
index 9aae2ccc7a..00698e30ef 100644
--- a/libavutil/cpu.h
+++ b/libavutil/cpu.h
@@ -82,6 +82,10 @@
 #define AV_CPU_FLAG_RVI          (1 << 0) ///< I (full GPR bank)
 #define AV_CPU_FLAG_RVF          (1 << 1) ///< F (single precision FP)
 #define AV_CPU_FLAG_RVD          (1 << 2) ///< D (double precision FP)
+#define AV_CPU_FLAG_RV_ZVE32X    (1 << 3) ///< Vectors of 8/16/32-bit int's */
+#define AV_CPU_FLAG_RV_ZVE32F    (1 << 4) ///< Vectors of float's */
+#define AV_CPU_FLAG_RV_ZVE64X    (1 << 5) ///< Vectors of 64-bit int's */
+#define AV_CPU_FLAG_RV_ZVE64D    (1 << 6) ///< Vectors of double's
 
 /**
  * Return the flags which specify extensions supported by the CPU.
diff --git a/libavutil/riscv/cpu.c b/libavutil/riscv/cpu.c
index fec1f7822a..6f862635b3 100644
--- a/libavutil/riscv/cpu.c
+++ b/libavutil/riscv/cpu.c
@@ -30,7 +30,32 @@
 
 int ff_force_cpu_flags_riscv(int flags)
 {
-    if ((flags & AV_CPU_FLAG_RVD) && !(flags & AV_CPU_FLAG_RVF)) {
+    if ((flags & AV_CPU_FLAG_RV_ZVE64D) && !(flags & AV_CPU_FLAG_RV_ZVE64X)) {
+        av_log(NULL, AV_LOG_WARNING, "RV%s implied by specified flags\n",
+               "_ZVE64X");
+        flags |= AV_CPU_FLAG_RV_ZVE64X;
+    }
+
+    if ((flags & AV_CPU_FLAG_RV_ZVE64D) && !(flags & AV_CPU_FLAG_RV_ZVE32F)) {
+        av_log(NULL, AV_LOG_WARNING, "RV%s implied by specified flags\n",
+               "_ZVE32F");
+        flags |= AV_CPU_FLAG_RV_ZVE32F;
+    }
+
+    if ((flags & (AV_CPU_FLAG_RV_ZVE64X | AV_CPU_FLAG_RV_ZVE32F))
+        && !(flags & AV_CPU_FLAG_RV_ZVE32X)) {
+        av_log(NULL, AV_LOG_WARNING, "RV%s implied by specified flags\n",
+               "_ZVE32X");
+        flags |= AV_CPU_FLAG_RV_ZVE32X;
+    }
+
+    if ((flags & AV_CPU_FLAG_RV_ZVE64D) && !(flags & AV_CPU_FLAG_RVD)) {
+        av_log(NULL, AV_LOG_WARNING, "RV%s implied by specified flags\n", "D");
+        flags |= AV_CPU_FLAG_RVD;
+    }
+
+    if ((flags & (AV_CPU_FLAG_RVD | AV_CPU_FLAG_RV_ZVE32F))
+        && !(flags & AV_CPU_FLAG_RVF)) {
         av_log(NULL, AV_LOG_WARNING, "RV%s implied by specified flags\n", "F");
         flags |= AV_CPU_FLAG_RVF;
     }
@@ -50,6 +75,11 @@ int ff_get_cpu_flags_riscv(void)
         ret |= AV_CPU_FLAG_RVF;
     if (hwcap & HWCAP_RV('D'))
         ret |= AV_CPU_FLAG_RVD;
+
+    /* The V extension implies all Zve* functional subsets */
+    if (hwcap & HWCAP_RV('V'))
+        ret |= AV_CPU_FLAG_RV_ZVE32X | AV_CPU_FLAG_RV_ZVE64X
+             | AV_CPU_FLAG_RV_ZVE32F | AV_CPU_FLAG_RV_ZVE64D;
 #endif
 
 #ifdef __riscv_i
@@ -60,6 +90,20 @@ int ff_get_cpu_flags_riscv(void)
 #if (__riscv_flen >= 64)
     ret |= AV_CPU_FLAG_RVD;
 #endif
+#endif
+
+    /* If RV-V is enabled statically at compile-time, check the details. */
+#ifdef __riscv_vectors
+    ret |= AV_CPU_FLAG_RV_ZVE32X;
+#if __riscv_v_elen >= 64
+    ret |= AV_CPU_FLAG_RV_ZVE64X;
+#endif
+#if __riscv_v_elen_fp >= 32
+    ret |= AV_CPU_FLAG_RV_ZVE32F;
+#if __riscv_v_elen_fp >= 64
+    ret |= AV_CPU_FLAG_RV_ZVE64F;
+#endif
+#endif
 #endif
 
     return ret;
diff --git a/tests/checkasm/checkasm.c b/tests/checkasm/checkasm.c
index 7730b14d98..c4352f1a16 100644
--- a/tests/checkasm/checkasm.c
+++ b/tests/checkasm/checkasm.c
@@ -230,9 +230,13 @@ static const struct {
     { "VSX",      "vsx",      AV_CPU_FLAG_VSX },
     { "POWER8",   "power8",   AV_CPU_FLAG_POWER8 },
 #elif ARCH_RISCV
-    { "RVI",      "rvi",      AV_CPU_FLAG_RVI },
-    { "RVF",      "rvf",      AV_CPU_FLAG_RVF },
-    { "RVD",      "rvd",      AV_CPU_FLAG_RVD },
+    { "RVI",        "rvi",       AV_CPU_FLAG_RVI },
+    { "RVF",        "rvf",       AV_CPU_FLAG_RVF },
+    { "RVD",        "rvd",       AV_CPU_FLAG_RVD },
+    { "RV_Zve32x",  "rv_zve32x", AV_CPU_FLAG_RV_ZVE32X },
+    { "RV_Zve32f",  "rv_zve32f", AV_CPU_FLAG_RV_ZVE32F },
+    { "RV_Zve64x",  "rv_zve64x", AV_CPU_FLAG_RV_ZVE64X },
+    { "RV_Zve64d",  "rv_zve64d", AV_CPU_FLAG_RV_ZVE64D },
 #elif ARCH_MIPS
     { "MMI",      "mmi",      AV_CPU_FLAG_MMI },
     { "MSA",      "msa",      AV_CPU_FLAG_MSA },
-- 
2.37.2

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

  parent reply	other threads:[~2022-09-20 14:42 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-20 14:39 [FFmpeg-devel] [PATCHv3 00/26] RISC-V CPU extensions Rémi Denis-Courmont
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 01/26] lavu/cpu: detect RISC-V base extensions remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 02/26] lavu/riscv: initial common header for assembler macros remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 03/26] lavc/audiodsp: RISC-V F vector_clipf remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 04/26] lavc/pixblockdsp: RISC-V I get_pixels remi
2022-09-20 14:39 ` remi [this message]
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 06/26] configure: probe RISC-V Vector extension remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 07/26] lavu/floatdsp: RISC-V V vector_fmul_scalar remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 08/26] lavu/floatdsp: RISC-V V vector_dmul_scalar remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 09/26] lavu/floatdsp: RISC-V V vector_fmul remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 10/26] lavu/floatdsp: RISC-V V vector_dmul remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 11/26] lavu/floatdsp: RISC-V V vector_fmac_scalar remi
2022-09-20 14:39 ` [FFmpeg-devel] [PATCH 12/26] lavu/floatdsp: RISC-V V vector_dmac_scalar remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 13/26] lavu/floatdsp: RISC-V V vector_fmul_add remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 14/26] lavu/floatdsp: RISC-V V butterflies_float remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 15/26] lavu/floatdsp: RISC-V V vector_fmul_reversed remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 16/26] lavu/floatdsp: RISC-V V vector_fmul_window remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 17/26] lavu/floatdsp: RISC-V V scalarproduct_float remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 18/26] lavu/fixeddsp: RISC-V V butterflies_fixed remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 19/26] lavc/audiodsp: RISC-V V vector_clip_int32 remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 20/26] lavc/audiodsp: RISC-V V vector_clipf remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 21/26] lavc/audiodsp: RISC-V V scalarproduct_int16 remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 22/26] lavc/fmtconvert: RISC-V V int32_to_float_fmul_scalar remi
2022-09-20 15:00   ` Andreas Rheinhardt
2022-09-20 15:12     ` Rémi Denis-Courmont
2022-09-21 14:14     ` [FFmpeg-devel] [PATCH 22 bis/26] " remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 23/26] lavc/fmtconvert: RISC-V V int32_to_float_fmul_array8 remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 24/26] lavc/vorbisdsp: RISC-V V inverse_coupling remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 25/26] lavc/aacpsdsp: RISC-V V add_squares remi
2022-09-20 14:40 ` [FFmpeg-devel] [PATCH 26/26] lavc/aacpsdsp: RISC-V V mul_pair_single remi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220920144013.4959-5-remi@remlab.net \
    --to=remi@remlab.net \
    --cc=ffmpeg-devel@ffmpeg.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
		ffmpegdev@gitmailbox.com
	public-inbox-index ffmpegdev

Example config snippet for mirrors.


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git