From patchwork Sun Jan 7 19:20:19 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Tobias Burnus X-Patchwork-Id: 83484 Return-Path: X-Original-To: patchwork@sourceware.org Delivered-To: patchwork@sourceware.org Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 2977F3861824 for ; Sun, 7 Jan 2024 19:20:59 +0000 (GMT) X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from mout.kundenserver.de (mout.kundenserver.de [212.227.126.131]) by sourceware.org (Postfix) with ESMTPS id 4056E3858C20 for ; Sun, 7 Jan 2024 19:20:24 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 4056E3858C20 Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=net-b.de Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=net-b.de ARC-Filter: OpenARC Filter v1.0.0 sourceware.org 4056E3858C20 Authentication-Results: server2.sourceware.org; arc=none smtp.remote-ip=212.227.126.131 ARC-Seal: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1704655229; cv=none; b=DilexJyhCdtDpNn3DUxhNvsnFNt2wxHKQ/+poeIpYIPtLYVXYRQjawwy4R0FLjrPfSCj62225bFB8hw4R4wqXx3lAriJmEBCAsOSItmhV6Hc3HS1dRJ4soModTUVjqLsX6rW/5Szbys+SxfX/i9SWQInS4DAgIfsH1QzLDVyAbs= ARC-Message-Signature: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1704655229; c=relaxed/simple; bh=nocfs5psuWe6sUxPAYR9N4fQVG8NTQf2x07I2/+W4w8=; h=Message-ID:Date:MIME-Version:To:From:Subject; b=lsj8oFOM/T3HaRjNPsYFMS8jvu6C5mynIjt/kFGV10YyDfeW4iR0EsWu4UOdG/LVfXRhLNCV2EnMivvCVzfJ2tlOe7LbZRrnex253Ne3Kgg+mRxqBIYBELrLhbH/+3SH59ox1I45WaCK3Y274/61dAApbvETod0AQ5zDskLtCjc= ARC-Authentication-Results: i=1; server2.sourceware.org Received: from [192.168.0.26] ([89.247.230.56]) by mrelayeu.kundenserver.de (mreue009 [213.165.67.97]) with ESMTPSA (Nemesis) id 1MRTEr-1rb38Y1WvF-00NP50; Sun, 07 Jan 2024 20:20:19 +0100 Message-ID: Date: Sun, 7 Jan 2024 20:20:19 +0100 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Content-Language: en-US To: gcc-patches , Andrew Stubbs From: Tobias Burnus Subject: [Patch] GCN: Add pre-initial support for gfx1100 X-Provags-ID: V03:K1:9eDzrCfOOFvUWV31Q0z3T8POdBip08kMHVkRLmmHMca7wGNoj+0 Est/zAS1y+KrRZBqMvzXZO63L8DKIVfXc6ttB4QYygfJzJim0rDSIELnJXFKG/B8sWc2a30 Kv1MF3+qIBeHnmXGTZtX94Mw9XdAGARwg9DyfTx7ra6j/JUykdiI4t/Kz93UYZ89uKKQs1s pjj4H/EOMuPIz+O/7gniQ== UI-OutboundReport: notjunk:1;M01:P0:tOvqx+YG0Ts=;6sD2pUAwU06sJxyw1uTdsXTRxH+ OaWSCC+ZcRooUrCrHDCmKx35zQPxjUXFGfc6umRUWIDrjVEDCPbQBAJzZRebJWCW79SVfrXJM 0hfdQw2feI4mknwz0Ns7oNWWCesTBdI0aYd8urHhOq4FEVx4f5y0+3XakUmR7UZsFNUllJ0du Q1VP5bvGP7rdodB3Iab7WcPSWK5jxhvzvAymOpjVMWUZYdaIMGGIr73yH/4Prj9On69sKXYs+ cP7wyVAxPuZfxXPP62JEXYRA/YwtRZPyvLSVKJf92W0f4PrMVdYTljmtUrNG3BE+xg5gDFOCo iNgb3FawvwhnqNVdSoPnweHx/ygHrH8sJpi8frEbUEi3FCL/v+ifSjY/k9xb85kPg9OWVKgDl fL3oZqS5zY4rIJWCOJxOyUiFmuWnOQtJaAA4VuZQTArMsTPeBaGXavVKs1O8DvTgfPIVXY9ZR tgRLIfLyCdFCvpuLH+qf8yL1ZtuQugHGWC6MbcBhgDRmB1O79JkUozNu9FsLIM/HHV9m3BCPE pWSgZz079ggIAPoYmKtiRIh8W45WTTnUOAFxxzj72A+3/bJxSJ54304lUbcnFk8+LrGIuKmSH ciejDVm+swQE7vAHCl9AwoY3hEN/1zE+q6HimVzwEVtLYA8X0WSbI6CCAjPYFMZnQGauQzlW7 hwC04fAKQXv6IKAZu9QmMiBXpIgYPiUWPew7HvFnqtoUozOukydc71orBFTKA/utGy84athRJ PKPwH3q7DQjP27yoAohdSQN5nbPnvEYcOyLble3on7L2/VxXIwkxhnphmZboqCyZvrCGifbKo 5gJMzVGlATVulRxFGFPyp/xCcgmw7Q+cG9i6LiWwZmJHJ/jEsg2/nchpLpTRdpIX7jdwvRXcj wjde6GPAI5pyW0Q== X-Spam-Status: No, score=-9.0 required=5.0 tests=BAYES_00, GIT_PATCH_0, KAM_DMARC_STATUS, KAM_NUMSUBJECT, RCVD_IN_BARRACUDACENTRAL, RCVD_IN_DNSWL_NONE, RCVD_IN_MSPIKE_H5, RCVD_IN_MSPIKE_WL, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: gcc-patches@gcc.gnu.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Gcc-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: gcc-patches-bounces+patchwork=sourceware.org@gcc.gnu.org ROCm meanwhile supports also some consumer cards; besides the semi-new gfx1030, support for gfx1100 was added more recently (in ROCm 5.7.1 for "Ubuntu 22.04 only" and without parenthesis since ROCm 6.0.0). GCC has already very limited support for gfx1030 - whose multlib support is - on purpose - not yet enabled by default and is WIP. The attached patch now adds gfx1100 on top of it, assuming that it mostly behaves the same as gfx1030. This is really WIP as there are known build (assembly) issues (see below) and not only "just" runtime issues. gfx1100 differs at least in the following aspects from the previously supported cards: * gfx1100 has an 'architected flat scratch' which is different from 'absolute flat scratch' which all others (but fiji: 'offset flat scratch') have. Hence, '.amdhsa_reserve_flat_scratch 0' has to be excluded to avoid assembly errors. * gfx1100 also does not support 'v_mov_b32_sdwa', failing to assembly libc/argz/libc_a-argz_stringify.o with: "sdwa variant of this instruction is not supported" → This has not been address in the patch, hence, specifying gfx1100 in --with-multilib-list= will fail to build when an in-tree newlib is build. * * * The attached patch fixes in addition one issue in libgomp (string-length len constant is too short for gfx1030 (and gfx1100) = 7 characters) and it includes the fix that __gfx1030__ is not defined, which I have submitted separately (yesterday). With the caveat that gfx1100 is even less usable than gfx1030 and it won't build newlib, is it nonetheless OK for mainline ? (As gfx1100 is not enabled by default in multilib, a regular build will will not fail and I think the *.md issue can be addressed separately.) Tobias GCN: Add pre-initial support for gfx1100 ROCm since 5.7.1 supports gfx1100 (RDNA3) cards. This commit adds support for it, mostly by assuming gfx1100 behaves identical to gfx1030. Like gfx1030, gfx1100 support is neither documented nor the build of the multilib enabled by default. But contrary to gfx1030, gfx1100 has a known issue causing some libraries not to build, including newlib: The sdwa variant of v_mov_b32_sdwa is not supported by the hardware but GCC current does generates this instruction. This will be addressed in a later commit. gcc/ChangeLog: * config.gcc (amdgcn-*-amdhsa): Accept --with-arch=gfx1100. * config/gcn/gcn-hsa.h (NO_XNACK): Add gfx1100: (ASM_SPEC): Handle gfx1100. * config/gcn/gcn-opts.h (enum processor_type): Add PROCESSOR_GFX1100. (enum gcn_isa): Add ISA_RDNA3. (TARGET_GFX1100, TARGET_RDNA2_PLUS, TARGET_RDNA3): Define. * config/gcn/gcn-valu.md: Change TARGET_RDNA2 to TARGET_RDNA2_PLUS. * config/gcn/gcn.cc (gcn_option_override, gcn_omp_device_kind_arch_isa, output_file_start): Handle gfx1100. (gcn_global_address_p, gcn_addr_space_legitimate_address_p): Change TARGET_RDNA2 to TARGET_RDNA2_PLUS. (gcn_hsa_declare_function_name): Don't use '.amdhsa_reserve_flat_scratch' with gfx1100. * config/gcn/gcn.h (ASSEMBLER_DIALECT): Likewise. (TARGET_CPU_CPP_BUILTINS): Define __RDNA3__, __gfx1030__ and __gfx1100__. * config/gcn/gcn.md: Change TARGET_RDNA2 to TARGET_RDNA2_PLUS. * config/gcn/gcn.opt (Enum gpu_type): Add gfx1100. * config/gcn/mkoffload.cc (EF_AMDGPU_MACH_AMDGCN_GFX1100): Define. (isa_has_combined_avgprs, main): Handle gfx1100. * config/gcn/t-omp-device (isa): Add gfx1100. libgomp/ChangeLog: * plugin/plugin-gcn.c (gcn_gfx1100_s): New const string. (gcn_isa_name_len): Fix length. (isa_hsa_name, isa_code, max_isa_vgprs): Handle gfx1100. gcc/config.gcc | 2 +- gcc/config/gcn/gcn-hsa.h | 4 ++-- gcc/config/gcn/gcn-opts.h | 7 +++++- gcc/config/gcn/gcn-valu.md | 10 ++++---- gcc/config/gcn/gcn.cc | 29 ++++++++++++++++------- gcc/config/gcn/gcn.h | 10 +++++--- gcc/config/gcn/gcn.md | 32 ++++++++++++------------- gcc/config/gcn/gcn.opt | 3 +++ gcc/config/gcn/mkoffload.cc | 5 ++++ gcc/config/gcn/t-omp-device | 2 +- gcc/tree-vect-loop-manip.cc | 16 +++++++++++++ gcc/tree-vect-loop.cc | 58 ++++++++++++++++++++++----------------------- libgomp/plugin/plugin-gcn.c | 9 ++++++- 13 files changed, 119 insertions(+), 68 deletions(-) diff --git a/gcc/config.gcc b/gcc/config.gcc index ce40b7758dd..7e583390024 100644 --- a/gcc/config.gcc +++ b/gcc/config.gcc @@ -4548,7 +4548,7 @@ case "${target}" in for which in arch tune; do eval "val=\$with_$which" case ${val} in - "" | fiji | gfx900 | gfx906 | gfx908 | gfx90a | gfx1030) + "" | fiji | gfx900 | gfx906 | gfx908 | gfx90a | gfx1030 | gfx1100) # OK ;; *) diff --git a/gcc/config/gcn/gcn-hsa.h b/gcc/config/gcn/gcn-hsa.h index 43bbe0411a3..bf7079fbbc6 100644 --- a/gcc/config/gcn/gcn-hsa.h +++ b/gcc/config/gcn/gcn-hsa.h @@ -75,7 +75,7 @@ extern unsigned int gcn_local_sym_hash (const char *name); supported for gcn. */ #define GOMP_SELF_SPECS "" -#define NO_XNACK "march=fiji:;march=gfx1030:;" \ +#define NO_XNACK "march=fiji:;march=gfx1030:;march=gfx1100:;" \ /* These match the defaults set in gcn.cc. */ \ "!mxnack*|mxnack=default:%{march=gfx900|march=gfx906|march=gfx908:-mattr=-xnack};" #define NO_SRAM_ECC "!march=*:;march=fiji:;march=gfx900:;march=gfx906:;" @@ -91,7 +91,7 @@ extern unsigned int gcn_local_sym_hash (const char *name); "%{!march=*|march=fiji:--amdhsa-code-object-version=3} " \ "%{" NO_XNACK XNACKOPT "}" \ "%{" NO_SRAM_ECC SRAMOPT "} " \ - "%{march=gfx1030:-mattr=+wavefrontsize64} " \ + "%{march=gfx1030|march=gfx1100:-mattr=+wavefrontsize64} " \ "-filetype=obj" #define LINK_SPEC "--pie --export-dynamic" #define LIB_SPEC "-lc" diff --git a/gcc/config/gcn/gcn-opts.h b/gcc/config/gcn/gcn-opts.h index 9a82cc81ea5..79fbda3ab25 100644 --- a/gcc/config/gcn/gcn-opts.h +++ b/gcc/config/gcn/gcn-opts.h @@ -25,7 +25,8 @@ enum processor_type PROCESSOR_VEGA20, // gfx906 PROCESSOR_GFX908, PROCESSOR_GFX90a, - PROCESSOR_GFX1030 + PROCESSOR_GFX1030, + PROCESSOR_GFX1100 }; #define TARGET_FIJI (gcn_arch == PROCESSOR_FIJI) @@ -34,6 +35,7 @@ enum processor_type #define TARGET_GFX908 (gcn_arch == PROCESSOR_GFX908) #define TARGET_GFX90a (gcn_arch == PROCESSOR_GFX90a) #define TARGET_GFX1030 (gcn_arch == PROCESSOR_GFX1030) +#define TARGET_GFX1100 (gcn_arch == PROCESSOR_GFX1100) /* Set in gcn_option_override. */ extern enum gcn_isa { @@ -41,6 +43,7 @@ extern enum gcn_isa { ISA_GCN3, ISA_GCN5, ISA_RDNA2, + ISA_RDNA3, ISA_CDNA1, ISA_CDNA2 } gcn_isa; @@ -54,6 +57,8 @@ extern enum gcn_isa { #define TARGET_CDNA2 (gcn_isa == ISA_CDNA2) #define TARGET_CDNA2_PLUS (gcn_isa >= ISA_CDNA2) #define TARGET_RDNA2 (gcn_isa == ISA_RDNA2) +#define TARGET_RDNA2_PLUS (gcn_isa >= ISA_RDNA2 && gcn_isa < ISA_CDNA1) +#define TARGET_RDNA3 (gcn_isa == ISA_RDNA3) #define TARGET_M0_LDS_LIMIT (TARGET_GCN3) diff --git a/gcc/config/gcn/gcn-valu.md b/gcc/config/gcn/gcn-valu.md index 615e48470dd..3d5b6271ee6 100644 --- a/gcc/config/gcn/gcn-valu.md +++ b/gcc/config/gcn/gcn-valu.md @@ -1417,7 +1417,7 @@ [(match_operand:V_noHI 1 "register_operand" " v") (match_operand:SI 2 "const_int_operand" " n")] UNSPEC_MOV_DPP_SHR))] - "!TARGET_RDNA2" + "!TARGET_RDNA2_PLUS" { return gcn_expand_dpp_shr_insn (mode, "v_mov_b32", UNSPEC_MOV_DPP_SHR, INTVAL (operands[2])); @@ -4211,7 +4211,7 @@ (unspec: [(match_operand:V_ALL 1 "register_operand")] REDUC_UNSPEC))] - "!TARGET_RDNA2" + "!TARGET_RDNA2_PLUS" { rtx tmp = gcn_expand_reduc_scalar (mode, operands[1], ); @@ -4265,7 +4265,7 @@ ; GCN3 requires a carry out, GCN5 not "!(TARGET_GCN3 && SCALAR_INT_MODE_P (mode) && == UNSPEC_PLUS_DPP_SHR) - && !TARGET_RDNA2" + && !TARGET_RDNA2_PLUS" { return gcn_expand_dpp_shr_insn (mode, "", , INTVAL (operands[3])); @@ -4310,7 +4310,7 @@ (match_operand:SI 3 "const_int_operand" "n")] UNSPEC_PLUS_CARRY_DPP_SHR)) (clobber (reg:DI VCC_REG))] - "!TARGET_RDNA2" + "!TARGET_RDNA2_PLUS" { return gcn_expand_dpp_shr_insn (mode, "v_add%^_u32", UNSPEC_PLUS_CARRY_DPP_SHR, @@ -4328,7 +4328,7 @@ (match_operand:DI 4 "register_operand" "cV")] UNSPEC_PLUS_CARRY_IN_DPP_SHR)) (clobber (reg:DI VCC_REG))] - "!TARGET_RDNA2" + "!TARGET_RDNA2_PLUS" { return gcn_expand_dpp_shr_insn (mode, "v_addc%^_u32", UNSPEC_PLUS_CARRY_IN_DPP_SHR, diff --git a/gcc/config/gcn/gcn.cc b/gcc/config/gcn/gcn.cc index 79d3a65ce13..50467bc6d39 100644 --- a/gcc/config/gcn/gcn.cc +++ b/gcc/config/gcn/gcn.cc @@ -139,6 +139,7 @@ gcn_option_override (void) : gcn_arch == PROCESSOR_GFX908 ? ISA_CDNA1 : gcn_arch == PROCESSOR_GFX90a ? ISA_CDNA2 : gcn_arch == PROCESSOR_GFX1030 ? ISA_RDNA2 + : gcn_arch == PROCESSOR_GFX1100 ? ISA_RDNA3 : ISA_UNKNOWN); gcc_assert (gcn_isa != ISA_UNKNOWN); @@ -160,15 +161,17 @@ gcn_option_override (void) acc_lds_size = 32768; } - /* gfx803 "Fiji" and gfx1030 do not support XNACK. */ + /* gfx803 "Fiji", gfx1030 and gfx1100 do not support XNACK. */ if (gcn_arch == PROCESSOR_FIJI - || gcn_arch == PROCESSOR_GFX1030) + || gcn_arch == PROCESSOR_GFX1030 + || gcn_arch == PROCESSOR_GFX1100) { if (flag_xnack == HSACO_ATTR_ON) - error ("-mxnack=on is incompatible with -march=%s", + error ("%<-mxnack=on%> is incompatible with %<-march=%s%>", (gcn_arch == PROCESSOR_FIJI ? "fiji" - : gcn_arch == PROCESSOR_GFX1030 ? "gfx1030" - : NULL)); + : gcn_arch == PROCESSOR_GFX1030 ? "gfx1030" + : gcn_arch == PROCESSOR_GFX1100 ? "gfx1100" + : NULL)); /* Allow HSACO_ATTR_ANY silently because that's the default. */ flag_xnack = HSACO_ATTR_OFF; } @@ -1592,7 +1595,7 @@ gcn_global_address_p (rtx addr) { rtx base = XEXP (addr, 0); rtx offset = XEXP (addr, 1); - int offsetbits = (TARGET_RDNA2 ? 11 : 12); + int offsetbits = (TARGET_RDNA2_PLUS ? 11 : 12); bool immediate_p = (CONST_INT_P (offset) && INTVAL (offset) >= -(1 << 12) && INTVAL (offset) < (1 << 12)); @@ -1725,7 +1728,7 @@ gcn_addr_space_legitimate_address_p (machine_mode mode, rtx x, bool strict, rtx base = XEXP (x, 0); rtx offset = XEXP (x, 1); - int offsetbits = (TARGET_RDNA2 ? 11 : 12); + int offsetbits = (TARGET_RDNA2_PLUS ? 11 : 12); bool immediate_p = (GET_CODE (offset) == CONST_INT /* Signed 12/13-bit immediate. */ && INTVAL (offset) >= -(1 << offsetbits) @@ -3043,6 +3046,8 @@ gcn_omp_device_kind_arch_isa (enum omp_device_kind_arch_isa trait, return gcn_arch == PROCESSOR_GFX90a; if (strcmp (name, "gfx1030") == 0) return gcn_arch == PROCESSOR_GFX1030; + if (strcmp (name, "gfx1100") == 0) + return gcn_arch == PROCESSOR_GFX1100; return 0; default: gcc_unreachable (); @@ -6539,6 +6544,11 @@ output_file_start (void) xnack = ""; sram_ecc = ""; break; + case PROCESSOR_GFX1100: + cpu = "gfx1100"; + xnack = ""; + sram_ecc = ""; + break; default: gcc_unreachable (); } @@ -6664,7 +6674,6 @@ gcn_hsa_declare_function_name (FILE *file, const char *name, tree decl) "\t .amdhsa_next_free_vgpr\t%i\n" "\t .amdhsa_next_free_sgpr\t%i\n" "\t .amdhsa_reserve_vcc\t1\n" - "\t .amdhsa_reserve_flat_scratch\t0\n" "\t .amdhsa_reserve_xnack_mask\t%i\n" "\t .amdhsa_private_segment_fixed_size\t0\n" "\t .amdhsa_group_segment_fixed_size\t%u\n" @@ -6674,6 +6683,10 @@ gcn_hsa_declare_function_name (FILE *file, const char *name, tree decl) sgpr, xnack_enabled, LDS_SIZE); + /* Not supported with 'architected flat scratch'. */ + if (gcn_arch != PROCESSOR_GFX1100) + fprintf (file, + "\t .amdhsa_reserve_flat_scratch\t0\n"); if (gcn_arch == PROCESSOR_GFX90a) fprintf (file, "\t .amdhsa_accum_offset\t%i\n" diff --git a/gcc/config/gcn/gcn.h b/gcc/config/gcn/gcn.h index c350cbb0545..c2afb5e9140 100644 --- a/gcc/config/gcn/gcn.h +++ b/gcc/config/gcn/gcn.h @@ -30,6 +30,8 @@ builtin_define ("__CDNA2__"); \ else if (TARGET_RDNA2) \ builtin_define ("__RDNA2__"); \ + else if (TARGET_RDNA3) \ + builtin_define ("__RDNA3__"); \ if (TARGET_FIJI) \ { \ builtin_define ("__fiji__"); \ @@ -41,11 +43,13 @@ builtin_define ("__gfx906__"); \ else if (TARGET_GFX908) \ builtin_define ("__gfx908__"); \ - else if (TARGET_GFX90a) \ - builtin_define ("__gfx90a__"); \ + else if (TARGET_GFX1030) \ + builtin_define ("__gfx1030"); \ + else if (TARGET_GFX1100) \ + builtin_define ("__gfx1100__"); \ } while (0) -#define ASSEMBLER_DIALECT (TARGET_RDNA2 ? 1 : 0) +#define ASSEMBLER_DIALECT (TARGET_RDNA2_PLUS ? 1 : 0) /* Support for a compile-time default architecture and tuning. The rules are: diff --git a/gcc/config/gcn/gcn.md b/gcc/config/gcn/gcn.md index e7815340002..492b833e255 100644 --- a/gcc/config/gcn/gcn.md +++ b/gcc/config/gcn/gcn.md @@ -299,10 +299,10 @@ (define_attr "enabled" "" (cond [(and (eq_attr "rdna" "no") - (ne (symbol_ref "TARGET_RDNA2") (const_int 0))) + (ne (symbol_ref "TARGET_RDNA2_PLUS") (const_int 0))) (const_int 0) (and (eq_attr "rdna" "yes") - (eq (symbol_ref "TARGET_RDNA2") (const_int 0))) + (eq (symbol_ref "TARGET_RDNA2_PLUS") (const_int 0))) (const_int 0) (and (eq_attr "gcn_version" "gcn5") (eq (symbol_ref "TARGET_GCN5_PLUS") (const_int 0))) @@ -2109,13 +2109,13 @@ return "s_load%o0\t%0, %A1 glc\;s_waitcnt\tlgkmcnt(0)\;" "s_dcache_wb_vol"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "flat_load%o0\t%0, %A1%O1 glc\;s_waitcnt\t0\;" "buffer_gl0_inv" : "flat_load%o0\t%0, %A1%O1 glc\;s_waitcnt\t0\;" "buffer_wbinvl1_vol"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "global_load%o0\t%0, %A1%O1 glc\;s_waitcnt\tvmcnt(0)\;" "buffer_gl0_inv" : "global_load%o0\t%0, %A1%O1 glc\;s_waitcnt\tvmcnt(0)\;" @@ -2131,13 +2131,13 @@ return "s_dcache_wb_vol\;s_load%o0\t%0, %A1 glc\;" "s_waitcnt\tlgkmcnt(0)\;s_dcache_inv_vol"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;flat_load%o0\t%0, %A1%O1 glc\;" "s_waitcnt\t0\;buffer_gl0_inv" : "buffer_wbinvl1_vol\;flat_load%o0\t%0, %A1%O1 glc\;" "s_waitcnt\t0\;buffer_wbinvl1_vol"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;global_load%o0\t%0, %A1%O1 glc\;" "s_waitcnt\tvmcnt(0)\;buffer_gl0_inv" : "buffer_wbinvl1_vol\;global_load%o0\t%0, %A1%O1 glc\;" @@ -2180,11 +2180,11 @@ case 0: return "s_dcache_wb_vol\;s_store%o1\t%1, %A0 glc"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;flat_store%o1\t%A0, %1%O0 glc" : "buffer_wbinvl1_vol\;flat_store%o1\t%A0, %1%O0 glc"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;global_store%o1\t%A0, %1%O0 glc" : "buffer_wbinvl1_vol\;global_store%o1\t%A0, %1%O0 glc"); } @@ -2198,13 +2198,13 @@ return "s_dcache_wb_vol\;s_store%o1\t%1, %A0 glc\;" "s_waitcnt\tlgkmcnt(0)\;s_dcache_inv_vol"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;flat_store%o1\t%A0, %1%O0 glc\;" "s_waitcnt\t0\;buffer_gl0_inv" : "buffer_wbinvl1_vol\;flat_store%o1\t%A0, %1%O0 glc\;" "s_waitcnt\t0\;buffer_wbinvl1_vol"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;global_store%o1\t%A0, %1%O0 glc\;" "s_waitcnt\tvmcnt(0)\;buffer_gl0_inv" : "buffer_wbinvl1_vol\;global_store%o1\t%A0, %1%O0 glc\;" @@ -2252,13 +2252,13 @@ return "s_atomic_swap\t%0, %1, %2 glc\;s_waitcnt\tlgkmcnt(0)\;" "s_dcache_wb_vol\;s_dcache_inv_vol"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "flat_atomic_swap\t%0, %1, %2 glc\;s_waitcnt\t0\;" "buffer_gl0_inv" : "flat_atomic_swap\t%0, %1, %2 glc\;s_waitcnt\t0\;" "buffer_wbinvl1_vol"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "global_atomic_swap\t%0, %A1, %2%O1 glc\;" "s_waitcnt\tvmcnt(0)\;buffer_gl0_inv" : "global_atomic_swap\t%0, %A1, %2%O1 glc\;" @@ -2273,13 +2273,13 @@ return "s_dcache_wb_vol\;s_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\tlgkmcnt(0)"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;flat_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\t0" : "buffer_wbinvl1_vol\;flat_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\t0"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;" "global_atomic_swap\t%0, %A1, %2%O1 glc\;" "s_waitcnt\tvmcnt(0)" @@ -2297,13 +2297,13 @@ return "s_dcache_wb_vol\;s_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\tlgkmcnt(0)\;s_dcache_inv_vol"; case 1: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;flat_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\t0\;buffer_gl0_inv" : "buffer_wbinvl1_vol\;flat_atomic_swap\t%0, %1, %2 glc\;" "s_waitcnt\t0\;buffer_wbinvl1_vol"); case 2: - return (TARGET_RDNA2 + return (TARGET_RDNA2_PLUS ? "buffer_gl0_inv\;" "global_atomic_swap\t%0, %A1, %2%O1 glc\;" "s_waitcnt\tvmcnt(0)\;buffer_gl0_inv" diff --git a/gcc/config/gcn/gcn.opt b/gcc/config/gcn/gcn.opt index b3d7a188d03..842fd36d25c 100644 --- a/gcc/config/gcn/gcn.opt +++ b/gcc/config/gcn/gcn.opt @@ -43,6 +43,9 @@ Enum(gpu_type) String(gfx90a) Value(PROCESSOR_GFX90a) EnumValue Enum(gpu_type) String(gfx1030) Value(PROCESSOR_GFX1030) +EnumValue +Enum(gpu_type) String(gfx1100) Value(PROCESSOR_GFX1100) + march= Target RejectNegative Negative(march=) Joined ToLower Enum(gpu_type) Var(gcn_arch) Init(PROCESSOR_FIJI) Specify the name of the target GPU. diff --git a/gcc/config/gcn/mkoffload.cc b/gcc/config/gcn/mkoffload.cc index 3341c0d34eb..5b680e6f13f 100644 --- a/gcc/config/gcn/mkoffload.cc +++ b/gcc/config/gcn/mkoffload.cc @@ -59,6 +59,8 @@ #define EF_AMDGPU_MACH_AMDGCN_GFX90a 0x3f #undef EF_AMDGPU_MACH_AMDGCN_GFX1030 #define EF_AMDGPU_MACH_AMDGCN_GFX1030 0x36 +#undef EF_AMDGPU_MACH_AMDGCN_GFX1100 +#define EF_AMDGPU_MACH_AMDGCN_GFX1100 0x41 #define EF_AMDGPU_FEATURE_XNACK_V4 0x300 /* Mask. */ #define EF_AMDGPU_FEATURE_XNACK_UNSUPPORTED_V4 0x000 @@ -484,6 +486,7 @@ isa_has_combined_avgprs (int isa) case EF_AMDGPU_MACH_AMDGCN_GFX906: case EF_AMDGPU_MACH_AMDGCN_GFX908: case EF_AMDGPU_MACH_AMDGCN_GFX1030: + case EF_AMDGPU_MACH_AMDGCN_GFX1100: return false; case EF_AMDGPU_MACH_AMDGCN_GFX90a: return true; @@ -1000,6 +1003,8 @@ main (int argc, char **argv) elf_arch = EF_AMDGPU_MACH_AMDGCN_GFX90a; else if (strcmp (argv[i], "-march=gfx1030") == 0) elf_arch = EF_AMDGPU_MACH_AMDGCN_GFX1030; + else if (strcmp (argv[i], "-march=gfx1100") == 0) + elf_arch = EF_AMDGPU_MACH_AMDGCN_GFX1100; #define STR "-mstack-size=" else if (startswith (argv[i], STR)) gcn_stack_size = atoi (argv[i] + strlen (STR)); diff --git a/gcc/config/gcn/t-omp-device b/gcc/config/gcn/t-omp-device index b1cd998a8b1..2315ad594a3 100644 --- a/gcc/config/gcn/t-omp-device +++ b/gcc/config/gcn/t-omp-device @@ -1,4 +1,4 @@ omp-device-properties-gcn: $(srcdir)/config/gcn/gcn.cc echo kind: gpu > $@ echo arch: amdgcn gcn >> $@ - echo isa: fiji gfx803 gfx900 gfx906 gfx908 gfx90a gfx1030 >> $@ + echo isa: fiji gfx803 gfx900 gfx906 gfx908 gfx90a gfx1030 gfx1100 >> $@ diff --git a/libgomp/plugin/plugin-gcn.c b/libgomp/plugin/plugin-gcn.c index bc8131a6c2d..e3e8b31c558 100644 --- a/libgomp/plugin/plugin-gcn.c +++ b/libgomp/plugin/plugin-gcn.c @@ -1641,7 +1641,8 @@ const static char *gcn_gfx906_s = "gfx906"; const static char *gcn_gfx908_s = "gfx908"; const static char *gcn_gfx90a_s = "gfx90a"; const static char *gcn_gfx1030_s = "gfx1030"; -const static int gcn_isa_name_len = 6; +const static char *gcn_gfx1100_s = "gfx1100"; +const static int gcn_isa_name_len = 7; /* Returns the name that the HSA runtime uses for the ISA or NULL if we do not support the ISA. */ @@ -1662,6 +1663,8 @@ isa_hsa_name (int isa) { return gcn_gfx90a_s; case EF_AMDGPU_MACH_AMDGCN_GFX1030: return gcn_gfx1030_s; + case EF_AMDGPU_MACH_AMDGCN_GFX1100: + return gcn_gfx1100_s; } return NULL; } @@ -1704,6 +1707,9 @@ isa_code(const char *isa) { if (!strncmp (isa, gcn_gfx1030_s, gcn_isa_name_len)) return EF_AMDGPU_MACH_AMDGCN_GFX1030; + if (!strncmp (isa, gcn_gfx1100_s, gcn_isa_name_len)) + return EF_AMDGPU_MACH_AMDGCN_GFX1100; + return -1; } @@ -1719,6 +1725,7 @@ max_isa_vgprs (int isa) case EF_AMDGPU_MACH_AMDGCN_GFX906: case EF_AMDGPU_MACH_AMDGCN_GFX908: case EF_AMDGPU_MACH_AMDGCN_GFX1030: + case EF_AMDGPU_MACH_AMDGCN_GFX1100: return 256; case EF_AMDGPU_MACH_AMDGCN_GFX90a: return 512;