mirror of https://github.com/gcc-mirror/gcc.git
amdgcn: Add gfx90c target
Add support for gfx90c GCN5 APU integrated graphics devices. The LLVM AMDGPU documentation does not list those devices as supported by rocm-amdhsa, but it passes most libgomp offloading tests. Although they are constrainted compared to dGPUs, they might be interesting for learning, experimentation, and testing. gcc/ChangeLog: * config.gcc: Add gfx90c. * config/gcn/gcn-hsa.h (NO_SRAM_ECC): Likewise. * config/gcn/gcn-opts.h (enum processor_type): Likewise. (TARGET_GFX90c): New macro. * config/gcn/gcn.cc (gcn_option_override): Handle gfx90c. (gcn_omp_device_kind_arch_isa): Likewise. (output_file_start): Likewise. * config/gcn/gcn.h: Add gfx90c. * config/gcn/gcn.opt: Likewise. * config/gcn/mkoffload.cc (EF_AMDGPU_MACH_AMDGCN_GFX90c): New macro. (get_arch): Handle gfx90c. (main): Handle EF_AMDGPU_MACH_AMDGCN_GFX90c * config/gcn/t-omp-device: Add gfx90c. * doc/install.texi: Likewise. * doc/invoke.texi: Likewise. libgomp/ChangeLog: * plugin/plugin-gcn.c (isa_hsa_name): Handle EF_AMDGPU_MACH_AMDGCN_GFX90c. (isa_code): Handle gfx90c. (max_isa_vgprs): Handle EF_AMDGPU_MACH_AMDGCN_GFX90c. Signed-off-by: Frederik Harwath <frederik@harwath.name>pull/62/merge
Frederik Harwath
parent
4a2e55b3ad
commit
b8e9fd535d
|
|
@ -4572,7 +4572,7 @@ case "${target}" in
|
|||
for which in arch tune; do
|
||||
eval "val=\$with_$which"
|
||||
case ${val} in
|
||||
"" | fiji | gfx900 | gfx906 | gfx908 | gfx90a | gfx1030 | gfx1036 | gfx1100 | gfx1103)
|
||||
"" | fiji | gfx900 | gfx906 | gfx908 | gfx90a | gfx90c | gfx1030 | gfx1036 | gfx1100 | gfx1103)
|
||||
# OK
|
||||
;;
|
||||
*)
|
||||
|
|
@ -4588,7 +4588,7 @@ case "${target}" in
|
|||
TM_MULTILIB_CONFIG=
|
||||
;;
|
||||
xdefault | xyes)
|
||||
TM_MULTILIB_CONFIG=`echo "gfx900,gfx906,gfx908,gfx90a,gfx1030,gfx1036,gfx1100,gfx1103" | sed "s/${with_arch},\?//;s/,$//"`
|
||||
TM_MULTILIB_CONFIG=`echo "gfx900,gfx906,gfx908,gfx90a,gfx90c,gfx1030,gfx1036,gfx1100,gfx1103" | sed "s/${with_arch},\?//;s/,$//"`
|
||||
;;
|
||||
*)
|
||||
TM_MULTILIB_CONFIG="${with_multilib_list}"
|
||||
|
|
|
|||
|
|
@ -93,7 +93,7 @@ extern unsigned int gcn_local_sym_hash (const char *name);
|
|||
#define NO_XNACK "march=fiji:;march=gfx1030:;march=gfx1036:;march=gfx1100:;march=gfx1103:;" \
|
||||
/* 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:;"
|
||||
#define NO_SRAM_ECC "!march=*:;march=fiji:;march=gfx900:;march=gfx906:;march=gfx90c:;"
|
||||
|
||||
/* In HSACOv4 no attribute setting means the binary supports "any" hardware
|
||||
configuration. The name of the attribute also changed. */
|
||||
|
|
|
|||
|
|
@ -25,6 +25,7 @@ enum processor_type
|
|||
PROCESSOR_VEGA20, // gfx906
|
||||
PROCESSOR_GFX908,
|
||||
PROCESSOR_GFX90a,
|
||||
PROCESSOR_GFX90c,
|
||||
PROCESSOR_GFX1030,
|
||||
PROCESSOR_GFX1036,
|
||||
PROCESSOR_GFX1100,
|
||||
|
|
@ -36,6 +37,7 @@ enum processor_type
|
|||
#define TARGET_VEGA20 (gcn_arch == PROCESSOR_VEGA20)
|
||||
#define TARGET_GFX908 (gcn_arch == PROCESSOR_GFX908)
|
||||
#define TARGET_GFX90a (gcn_arch == PROCESSOR_GFX90a)
|
||||
#define TARGET_GFX90c (gcn_arch == PROCESSOR_GFX90c)
|
||||
#define TARGET_GFX1030 (gcn_arch == PROCESSOR_GFX1030)
|
||||
#define TARGET_GFX1036 (gcn_arch == PROCESSOR_GFX1036)
|
||||
#define TARGET_GFX1100 (gcn_arch == PROCESSOR_GFX1100)
|
||||
|
|
|
|||
|
|
@ -138,6 +138,7 @@ gcn_option_override (void)
|
|||
: gcn_arch == PROCESSOR_VEGA20 ? ISA_GCN5
|
||||
: gcn_arch == PROCESSOR_GFX908 ? ISA_CDNA1
|
||||
: gcn_arch == PROCESSOR_GFX90a ? ISA_CDNA2
|
||||
: gcn_arch == PROCESSOR_GFX90c ? ISA_GCN5
|
||||
: gcn_arch == PROCESSOR_GFX1030 ? ISA_RDNA2
|
||||
: gcn_arch == PROCESSOR_GFX1036 ? ISA_RDNA2
|
||||
: gcn_arch == PROCESSOR_GFX1100 ? ISA_RDNA3
|
||||
|
|
@ -196,6 +197,7 @@ gcn_option_override (void)
|
|||
flag_xnack = HSACO_ATTR_OFF;
|
||||
break;
|
||||
case PROCESSOR_GFX90a:
|
||||
case PROCESSOR_GFX90c:
|
||||
flag_xnack = HSACO_ATTR_ANY;
|
||||
break;
|
||||
default:
|
||||
|
|
@ -3050,6 +3052,8 @@ gcn_omp_device_kind_arch_isa (enum omp_device_kind_arch_isa trait,
|
|||
return gcn_arch == PROCESSOR_GFX908;
|
||||
if (strcmp (name, "gfx90a") == 0)
|
||||
return gcn_arch == PROCESSOR_GFX90a;
|
||||
if (strcmp (name, "gfx90c") == 0)
|
||||
return gcn_arch == PROCESSOR_GFX90c;
|
||||
if (strcmp (name, "gfx1030") == 0)
|
||||
return gcn_arch == PROCESSOR_GFX1030;
|
||||
if (strcmp (name, "gfx1036") == 0)
|
||||
|
|
@ -6596,6 +6600,10 @@ output_file_start (void)
|
|||
case PROCESSOR_GFX90a:
|
||||
cpu = "gfx90a";
|
||||
break;
|
||||
case PROCESSOR_GFX90c:
|
||||
cpu = "gfx90c";
|
||||
sram_ecc = "";
|
||||
break;
|
||||
case PROCESSOR_GFX1030:
|
||||
cpu = "gfx1030";
|
||||
xnack = "";
|
||||
|
|
|
|||
|
|
@ -47,6 +47,8 @@
|
|||
builtin_define ("__gfx908__"); \
|
||||
else if (TARGET_GFX90a) \
|
||||
builtin_define ("__gfx90a__"); \
|
||||
else if (TARGET_GFX90c) \
|
||||
builtin_define ("__gfx90c__"); \
|
||||
else if (TARGET_GFX1030) \
|
||||
builtin_define ("__gfx1030__"); \
|
||||
else if (TARGET_GFX1036) \
|
||||
|
|
|
|||
|
|
@ -40,6 +40,9 @@ Enum(gpu_type) String(gfx908) Value(PROCESSOR_GFX908)
|
|||
EnumValue
|
||||
Enum(gpu_type) String(gfx90a) Value(PROCESSOR_GFX90a)
|
||||
|
||||
EnumValue
|
||||
Enum(gpu_type) String(gfx90c) Value(PROCESSOR_GFX90c)
|
||||
|
||||
EnumValue
|
||||
Enum(gpu_type) String(gfx1030) Value(PROCESSOR_GFX1030)
|
||||
|
||||
|
|
|
|||
|
|
@ -59,6 +59,8 @@
|
|||
#define EF_AMDGPU_MACH_AMDGCN_GFX908 0x30
|
||||
#undef EF_AMDGPU_MACH_AMDGCN_GFX90a
|
||||
#define EF_AMDGPU_MACH_AMDGCN_GFX90a 0x3f
|
||||
#undef EF_AMDGPU_MACH_AMDGCN_GFX90c
|
||||
#define EF_AMDGPU_MACH_AMDGCN_GFX90c 0x32
|
||||
#undef EF_AMDGPU_MACH_AMDGCN_GFX1030
|
||||
#define EF_AMDGPU_MACH_AMDGCN_GFX1030 0x36
|
||||
#undef EF_AMDGPU_MACH_AMDGCN_GFX1036
|
||||
|
|
@ -861,6 +863,8 @@ get_arch (const char *str, const char *with_arch_str)
|
|||
return EF_AMDGPU_MACH_AMDGCN_GFX908;
|
||||
else if (strcmp (str, "gfx90a") == 0)
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX90a;
|
||||
else if (strcmp (str, "gfx90c") == 0)
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX90c;
|
||||
else if (strcmp (str, "gfx1030") == 0)
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX1030;
|
||||
else if (strcmp (str, "gfx1036") == 0)
|
||||
|
|
@ -1099,6 +1103,11 @@ main (int argc, char **argv)
|
|||
if (TEST_SRAM_ECC_UNSET (elf_flags))
|
||||
SET_SRAM_ECC_ANY (elf_flags);
|
||||
break;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX90c:
|
||||
if (TEST_XNACK_UNSET (elf_flags))
|
||||
SET_XNACK_ANY (elf_flags);
|
||||
SET_SRAM_ECC_UNSET (elf_flags);
|
||||
break;
|
||||
default:
|
||||
fatal_error (input_location, "unhandled architecture");
|
||||
}
|
||||
|
|
|
|||
|
|
@ -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 gfx1036 gfx1100 gfx1103 >> $@
|
||||
echo isa: fiji gfx803 gfx900 gfx906 gfx908 gfx90a gfx90c gfx1030 gfx1036 gfx1100 gfx1103 >> $@
|
||||
|
|
|
|||
|
|
@ -1270,8 +1270,8 @@ default set of libraries is selected based on the value of
|
|||
|
||||
@item amdgcn*-*-*
|
||||
@var{list} is a comma separated list of ISA names (allowed values: @code{fiji},
|
||||
@code{gfx900}, @code{gfx906}, @code{gfx908}, @code{gfx90a}, @code{gfx1030},
|
||||
@code{gfx1036}, @code{gfx1100}, @code{gfx1103}).
|
||||
@code{gfx900}, @code{gfx906}, @code{gfx908}, @code{gfx90a}, @code{gfx90c},
|
||||
@code{gfx1030}, @code{gfx1036}, @code{gfx1100}, @code{gfx1103}).
|
||||
It ought not include the name of the default
|
||||
ISA, specified via @option{--with-arch}. If @var{list} is empty, then there
|
||||
will be no multilibs and only the default run-time library will be built. If
|
||||
|
|
|
|||
|
|
@ -21844,6 +21844,9 @@ Compile for CDNA1 Instinct MI100 series devices (gfx908).
|
|||
@item gfx90a
|
||||
Compile for CDNA2 Instinct MI200 series devices (gfx90a).
|
||||
|
||||
@item gfx90c
|
||||
Compile for GCN5 Vega 7 devices (gfx90c).
|
||||
|
||||
@item gfx1030
|
||||
Compile for RDNA2 gfx1030 devices (GFX10 series).
|
||||
|
||||
|
|
|
|||
|
|
@ -390,6 +390,7 @@ typedef enum {
|
|||
EF_AMDGPU_MACH_AMDGCN_GFX906 = 0x02f,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX908 = 0x030,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX90a = 0x03f,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX90c = 0x032,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX1030 = 0x036,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX1036 = 0x045,
|
||||
EF_AMDGPU_MACH_AMDGCN_GFX1100 = 0x041,
|
||||
|
|
@ -1679,6 +1680,7 @@ const static char *gcn_gfx900_s = "gfx900";
|
|||
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_gfx90c_s = "gfx90c";
|
||||
const static char *gcn_gfx1030_s = "gfx1030";
|
||||
const static char *gcn_gfx1036_s = "gfx1036";
|
||||
const static char *gcn_gfx1100_s = "gfx1100";
|
||||
|
|
@ -1702,6 +1704,8 @@ isa_hsa_name (int isa) {
|
|||
return gcn_gfx908_s;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX90a:
|
||||
return gcn_gfx90a_s;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX90c:
|
||||
return gcn_gfx90c_s;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX1030:
|
||||
return gcn_gfx1030_s;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX1036:
|
||||
|
|
@ -1749,6 +1753,9 @@ isa_code(const char *isa) {
|
|||
if (!strncmp (isa, gcn_gfx90a_s, gcn_isa_name_len))
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX90a;
|
||||
|
||||
if (!strncmp (isa, gcn_gfx90c_s, gcn_isa_name_len))
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX90c;
|
||||
|
||||
if (!strncmp (isa, gcn_gfx1030_s, gcn_isa_name_len))
|
||||
return EF_AMDGPU_MACH_AMDGCN_GFX1030;
|
||||
|
||||
|
|
@ -1778,6 +1785,8 @@ max_isa_vgprs (int isa)
|
|||
return 256;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX90a:
|
||||
return 512;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX90c:
|
||||
return 256;
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX1030:
|
||||
case EF_AMDGPU_MACH_AMDGCN_GFX1036:
|
||||
return 512; /* 512 SIMD32 = 256 wavefrontsize64. */
|
||||
|
|
|
|||
Loading…
Reference in New Issue