Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
composable_kernel_ROCM
Commits
c7a9ad98
Commit
c7a9ad98
authored
Feb 07, 2024
by
Adam Osewski
Browse files
Fix Navi compilation.
parent
e878371c
Changes
2
Show whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
9 additions
and
0 deletions
+9
-0
include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_multiple_d_splitk_xdl_cshuffle_tile_loop.hpp
...grouped_gemm_multiple_d_splitk_xdl_cshuffle_tile_loop.hpp
+3
-0
library/include/ck/library/tensor_operation_instance/gpu/grouped_gemm_multiple_d.hpp
...tensor_operation_instance/gpu/grouped_gemm_multiple_d.hpp
+6
-0
No files found.
include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_multiple_d_splitk_xdl_cshuffle_tile_loop.hpp
View file @
c7a9ad98
...
@@ -215,6 +215,9 @@ __global__ void
...
@@ -215,6 +215,9 @@ __global__ void
ignore
=
p_workspace
;
ignore
=
p_workspace
;
ignore
=
tile_count
;
ignore
=
tile_count
;
ignore
=
k_batch
;
ignore
=
k_batch
;
ignore
=
a_element_op
;
ignore
=
b_element_op
;
ignore
=
cde_element_op
;
#endif // end of if (defined(__gfx908__) || defined(__gfx90a__))
#endif // end of if (defined(__gfx908__) || defined(__gfx90a__))
}
}
...
...
library/include/ck/library/tensor_operation_instance/gpu/grouped_gemm_multiple_d.hpp
View file @
c7a9ad98
...
@@ -16,6 +16,7 @@ namespace tensor_operation {
...
@@ -16,6 +16,7 @@ namespace tensor_operation {
namespace
device
{
namespace
device
{
namespace
instance
{
namespace
instance
{
#if defined(CK_ENABLE_FP16)
// MultiD version
// MultiD version
void
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_nk_mn_irregular_instances
(
void
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_nk_mn_irregular_instances
(
std
::
vector
<
std
::
unique_ptr
<
DeviceGroupedGemm
<
Row
,
std
::
vector
<
std
::
unique_ptr
<
DeviceGroupedGemm
<
Row
,
...
@@ -42,6 +43,7 @@ void add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_kn_mn_irregu
...
@@ -42,6 +43,7 @@ void add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_kn_mn_irregu
PassThrough
,
PassThrough
,
PassThrough
,
PassThrough
,
PassThrough
>>>&
instances
);
PassThrough
>>>&
instances
);
#endif
template
<
typename
ALayout
,
template
<
typename
ALayout
,
typename
BLayout
,
typename
BLayout
,
...
@@ -87,14 +89,18 @@ struct DeviceOperationInstanceFactory<
...
@@ -87,14 +89,18 @@ struct DeviceOperationInstanceFactory<
if
constexpr
(
is_same_v
<
ALayout
,
Row
>
&&
is_same_v
<
BLayout
,
Row
>
&&
if
constexpr
(
is_same_v
<
ALayout
,
Row
>
&&
is_same_v
<
BLayout
,
Row
>
&&
is_same_v
<
ELayout
,
Row
>
)
is_same_v
<
ELayout
,
Row
>
)
{
{
#if defined(CK_ENABLE_FP16)
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_kn_mn_irregular_instances
(
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_kn_mn_irregular_instances
(
op_ptrs
);
op_ptrs
);
#endif
}
}
else
if
constexpr
(
is_same_v
<
ALayout
,
Row
>
&&
is_same_v
<
BLayout
,
Col
>
&&
else
if
constexpr
(
is_same_v
<
ALayout
,
Row
>
&&
is_same_v
<
BLayout
,
Col
>
&&
is_same_v
<
ELayout
,
Row
>
)
is_same_v
<
ELayout
,
Row
>
)
{
{
#if defined(CK_ENABLE_FP16)
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_nk_mn_irregular_instances
(
add_device_grouped_gemm_multi_d_splitk_cshuffle_f16_f16_f16_mk_nk_mn_irregular_instances
(
op_ptrs
);
op_ptrs
);
#endif
}
}
}
}
return
op_ptrs
;
return
op_ptrs
;
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment