From: remi@remlab.net To: ffmpeg-devel@ffmpeg.org Subject: [FFmpeg-devel] [PATCH 05/31] lavu/cpu: CPU flags for the RISC-V Vector extension Date: Mon, 26 Sep 2022 17:52:25 +0300 Message-ID: <20220926145251.56351-5-remi@remlab.net> (raw) In-Reply-To: <5862173.lOV4Wx5bFT@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 (RVV_I32), floats (RVV_F32), 64-bit ints (RVV_I64) and doubles (RVV_F64). 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 | 19 +++++++++++++++++++ tests/checkasm/checkasm.c | 4 ++++ 4 files changed, 31 insertions(+) diff --git a/libavutil/cpu.c b/libavutil/cpu.c index 8b6eef9873..5818fd9c1c 100644 --- a/libavutil/cpu.c +++ b/libavutil/cpu.c @@ -184,6 +184,10 @@ 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" }, + { "rvv-i32", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVV_I32 }, .unit = "flags" }, + { "rvv-f32", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVV_F32 }, .unit = "flags" }, + { "rvv-i64", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVV_I64 }, .unit = "flags" }, + { "rvv", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = AV_CPU_FLAG_RVV_F64 }, .unit = "flags" }, #endif { NULL }, }; diff --git a/libavutil/cpu.h b/libavutil/cpu.h index 9aae2ccc7a..18f42af015 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_RVV_I32 (1 << 3) ///< Vectors of 8/16/32-bit int's */ +#define AV_CPU_FLAG_RVV_F32 (1 << 4) ///< Vectors of float's */ +#define AV_CPU_FLAG_RVV_I64 (1 << 5) ///< Vectors of 64-bit int's */ +#define AV_CPU_FLAG_RVV_F64 (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 6803f035e5..e234201395 100644 --- a/libavutil/riscv/cpu.c +++ b/libavutil/riscv/cpu.c @@ -40,6 +40,11 @@ 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_RVV_I32 | AV_CPU_FLAG_RVV_I64 + | AV_CPU_FLAG_RVV_F32 | AV_CPU_FLAG_RVV_F64; #endif #ifdef __riscv_i @@ -50,6 +55,20 @@ 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_RVV_I32; +#if __riscv_v_elen >= 64 + ret |= AV_CPU_FLAG_RVV_I64; +#endif +#if __riscv_v_elen_fp >= 32 + ret |= AV_CPU_FLAG_RVV_F32; +#if __riscv_v_elen_fp >= 64 + ret |= AV_CPU_FLAG_RVV_F64; +#endif +#endif #endif return ret; diff --git a/tests/checkasm/checkasm.c b/tests/checkasm/checkasm.c index e1135a84ac..90dd7e4634 100644 --- a/tests/checkasm/checkasm.c +++ b/tests/checkasm/checkasm.c @@ -236,6 +236,10 @@ static const struct { { "RVI", "rvi", AV_CPU_FLAG_RVI }, { "RVF", "rvf", AV_CPU_FLAG_RVF }, { "RVD", "rvd", AV_CPU_FLAG_RVD }, + { "RVVi32", "rvv_i32", AV_CPU_FLAG_RVV_I32 }, + { "RVVf32", "rvv_f32", AV_CPU_FLAG_RVV_F32 }, + { "RVVi64", "rvv_i64", AV_CPU_FLAG_RVV_I64 }, + { "RVVf64", "rvv_f64", AV_CPU_FLAG_RVV_F64 }, #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".
next prev parent reply other threads:[~2022-09-26 14:54 UTC|newest] Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-09-26 14:52 [FFmpeg-devel] [PATCHv6 00/31] initial RISC-V CPU extensions Rémi Denis-Courmont 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 01/31] lavu/cpu: detect RISC-V base extensions remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 02/31] lavu/riscv: initial common header for assembler macros remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 03/31] lavc/audiodsp: RISC-V F vector_clipf remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 04/31] lavc/pixblockdsp: RISC-V I get_pixels remi 2022-09-26 14:52 ` remi [this message] 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 06/31] configure: probe RISC-V Vector extension remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 07/31] lavu/riscv: fallback macros for SH{1, 2, 3}ADD remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 08/31] lavu/floatdsp: RISC-V V vector_fmul_scalar remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 09/31] lavu/floatdsp: RISC-V V vector_dmul_scalar remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 10/31] lavu/floatdsp: RISC-V V vector_fmul remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 11/31] lavu/floatdsp: RISC-V V vector_dmul remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 12/31] lavu/floatdsp: RISC-V V vector_fmac_scalar remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 13/31] lavu/floatdsp: RISC-V V vector_dmac_scalar remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 14/31] lavu/floatdsp: RISC-V V vector_fmul_add remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 15/31] lavu/floatdsp: RISC-V V butterflies_float remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 16/31] lavu/floatdsp: RISC-V V vector_fmul_reverse remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 17/31] lavu/floatdsp: RISC-V V vector_fmul_window remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 18/31] lavu/floatdsp: RISC-V V scalarproduct_float remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 19/31] lavu/fixeddsp: RISC-V V butterflies_fixed remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 20/31] lavc/audiodsp: RISC-V V vector_clip_int32 remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 21/31] lavc/audiodsp: RISC-V V vector_clipf remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 22/31] lavc/audiodsp: RISC-V V scalarproduct_int16 remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 23/31] lavc/fmtconvert: RISC-V V int32_to_float_fmul_scalar remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 24/31] lavc/fmtconvert: RISC-V V int32_to_float_fmul_array8 remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 25/31] lavc/vorbisdsp: RISC-V V inverse_coupling remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 26/31] lavc/aacpsdsp: RISC-V V add_squares remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 27/31] lavc/aacpsdsp: RISC-V V mul_pair_single remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 28/31] lavc/aacpsdsp: RISC-V V hybrid_analysis remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 29/31] lavc/aacpsdsp: RISC-V V hybrid_analysis_ileave remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 30/31] lavc/aacpsdsp: RISC-V V hybrid_synthesis_deint remi 2022-09-26 14:52 ` [FFmpeg-devel] [PATCH 31/31] lavc/aacpsdsp: RISC-V V stereo_interpolate[0] remi 2022-09-27 12:01 ` [FFmpeg-devel] [PATCHv6 00/31] initial RISC-V CPU extensions Lynne -- strict thread matches above, loose matches on Subject: below -- 2022-09-25 14:25 [FFmpeg-devel] [PATCHv5 00/31] " Rémi Denis-Courmont 2022-09-25 14:25 ` [FFmpeg-devel] [PATCH 05/31] lavu/cpu: CPU flags for the RISC-V Vector extension remi 2022-09-26 6:51 ` Lynne 2022-09-26 8:02 ` Andreas Rheinhardt 2022-09-26 9:38 ` Rémi Denis-Courmont
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=20220926145251.56351-5-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