* [PATCH] mldev: remove weak symbols use in type conversions
@ 2023-03-11 14:50 Srikanth Yalavarthi
2023-03-13 10:52 ` David Marchand
2023-03-13 11:36 ` [PATCH v2 1/1] " Srikanth Yalavarthi
0 siblings, 2 replies; 5+ messages in thread
From: Srikanth Yalavarthi @ 2023-03-11 14:50 UTC (permalink / raw)
To: Srikanth Yalavarthi, David Marchand; +Cc: dev, sshankarnara
Drop use of weak symbols to select the type conversion
functions. Select NEON implementation only on Aarch64
builds. Enable windows build.
Fixes: 3c54f91ad8c9 ("mldev: disable build on Windows")
Signed-off-by: Srikanth Yalavarthi <syalavarthi@marvell.com>
---
lib/mldev/meson.build | 9 ++-------
lib/mldev/mldev_utils_scalar.c | 24 ++++++++++++------------
2 files changed, 14 insertions(+), 19 deletions(-)
diff --git a/lib/mldev/meson.build b/lib/mldev/meson.build
index 51ba809465..c9db42257b 100644
--- a/lib/mldev/meson.build
+++ b/lib/mldev/meson.build
@@ -1,21 +1,16 @@
# SPDX-License-Identifier: BSD-3-Clause
# Copyright (c) 2022 Marvell.
-if is_windows
- build = false
- reason = 'not supported on Windows'
- subdir_done()
-endif
-
sources = files(
'rte_mldev_pmd.c',
'rte_mldev.c',
'mldev_utils.c',
- 'mldev_utils_scalar.c',
)
if dpdk_conf.has('RTE_ARCH_ARM64')
sources += files('mldev_utils_neon.c')
+else
+ sources += files('mldev_utils_scalar.c')
endif
headers = files(
diff --git a/lib/mldev/mldev_utils_scalar.c b/lib/mldev/mldev_utils_scalar.c
index 40320ed3ef..322b009f5d 100644
--- a/lib/mldev/mldev_utils_scalar.c
+++ b/lib/mldev/mldev_utils_scalar.c
@@ -84,7 +84,7 @@ union float32 {
uint32_t u;
};
-__rte_weak int
+int
rte_ml_io_float32_to_int8(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -116,7 +116,7 @@ rte_ml_io_float32_to_int8(float scale, uint64_t nb_elements, void *input, void *
return 0;
}
-__rte_weak int
+int
rte_ml_io_int8_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
int8_t *input_buffer;
@@ -139,7 +139,7 @@ rte_ml_io_int8_to_float32(float scale, uint64_t nb_elements, void *input, void *
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_uint8(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -171,7 +171,7 @@ rte_ml_io_float32_to_uint8(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_uint8_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
uint8_t *input_buffer;
@@ -194,7 +194,7 @@ rte_ml_io_uint8_to_float32(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_int16(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -226,7 +226,7 @@ rte_ml_io_float32_to_int16(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_int16_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
int16_t *input_buffer;
@@ -249,7 +249,7 @@ rte_ml_io_int16_to_float32(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_uint16(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -281,7 +281,7 @@ rte_ml_io_float32_to_uint16(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_uint16_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
@@ -429,7 +429,7 @@ __float32_to_float16_scalar_rtn(float x)
return u16;
}
-__rte_weak int
+int
rte_ml_io_float32_to_float16(uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -513,7 +513,7 @@ __float16_to_float32_scalar_rtx(uint16_t f16)
return f32.f;
}
-__rte_weak int
+int
rte_ml_io_float16_to_float32(uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
@@ -614,7 +614,7 @@ __float32_to_bfloat16_scalar_rtn(float x)
return u16;
}
-__rte_weak int
+int
rte_ml_io_float32_to_bfloat16(uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -696,7 +696,7 @@ __bfloat16_to_float32_scalar_rtx(uint16_t f16)
return f32.f;
}
-__rte_weak int
+int
rte_ml_io_bfloat16_to_float32(uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
--
2.17.1
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] mldev: remove weak symbols use in type conversions
2023-03-11 14:50 [PATCH] mldev: remove weak symbols use in type conversions Srikanth Yalavarthi
@ 2023-03-13 10:52 ` David Marchand
2023-03-13 11:37 ` [EXT] " Srikanth Yalavarthi
2023-03-13 11:36 ` [PATCH v2 1/1] " Srikanth Yalavarthi
1 sibling, 1 reply; 5+ messages in thread
From: David Marchand @ 2023-03-13 10:52 UTC (permalink / raw)
To: Srikanth Yalavarthi; +Cc: dev, sshankarnara
Hello,
On Sat, Mar 11, 2023 at 3:51 PM Srikanth Yalavarthi
<syalavarthi@marvell.com> wrote:
>
> Drop use of weak symbols to select the type conversion
> functions. Select NEON implementation only on Aarch64
> builds. Enable windows build.
>
> Fixes: 3c54f91ad8c9 ("mldev: disable build on Windows")
Strictly speaking we are not fixing this change above, but rather
Fixes: 9637de38a2e3 ("mldev: add scalar type conversion")
>
> Signed-off-by: Srikanth Yalavarthi <syalavarthi@marvell.com>
The change looks good to me and the UNH Windows build target seems to agree :-).
Reviewed-by: David Marchand <david.marchand@redhat.com>
Thanks Srikanth.
--
David Marchand
^ permalink raw reply [flat|nested] 5+ messages in thread
* [PATCH v2 1/1] mldev: remove weak symbols use in type conversions
2023-03-11 14:50 [PATCH] mldev: remove weak symbols use in type conversions Srikanth Yalavarthi
2023-03-13 10:52 ` David Marchand
@ 2023-03-13 11:36 ` Srikanth Yalavarthi
2023-03-15 9:49 ` David Marchand
1 sibling, 1 reply; 5+ messages in thread
From: Srikanth Yalavarthi @ 2023-03-13 11:36 UTC (permalink / raw)
To: Srikanth Yalavarthi
Cc: dev, sshankarnara, aprabhu, ptakkar, pshukla, david.marchand
Drop use of weak symbols to select the type conversion
functions. Select NEON implementation only on Aarch64
builds. Enable windows build.
Fixes: 9637de38a2e3 ("mldev: add scalar type conversion")
Signed-off-by: Srikanth Yalavarthi <syalavarthi@marvell.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
---
v2:
* Updated Fixes reference
lib/mldev/meson.build | 9 ++-------
lib/mldev/mldev_utils_scalar.c | 24 ++++++++++++------------
2 files changed, 14 insertions(+), 19 deletions(-)
diff --git a/lib/mldev/meson.build b/lib/mldev/meson.build
index 51ba809465..c9db42257b 100644
--- a/lib/mldev/meson.build
+++ b/lib/mldev/meson.build
@@ -1,21 +1,16 @@
# SPDX-License-Identifier: BSD-3-Clause
# Copyright (c) 2022 Marvell.
-if is_windows
- build = false
- reason = 'not supported on Windows'
- subdir_done()
-endif
-
sources = files(
'rte_mldev_pmd.c',
'rte_mldev.c',
'mldev_utils.c',
- 'mldev_utils_scalar.c',
)
if dpdk_conf.has('RTE_ARCH_ARM64')
sources += files('mldev_utils_neon.c')
+else
+ sources += files('mldev_utils_scalar.c')
endif
headers = files(
diff --git a/lib/mldev/mldev_utils_scalar.c b/lib/mldev/mldev_utils_scalar.c
index 40320ed3ef..322b009f5d 100644
--- a/lib/mldev/mldev_utils_scalar.c
+++ b/lib/mldev/mldev_utils_scalar.c
@@ -84,7 +84,7 @@ union float32 {
uint32_t u;
};
-__rte_weak int
+int
rte_ml_io_float32_to_int8(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -116,7 +116,7 @@ rte_ml_io_float32_to_int8(float scale, uint64_t nb_elements, void *input, void *
return 0;
}
-__rte_weak int
+int
rte_ml_io_int8_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
int8_t *input_buffer;
@@ -139,7 +139,7 @@ rte_ml_io_int8_to_float32(float scale, uint64_t nb_elements, void *input, void *
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_uint8(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -171,7 +171,7 @@ rte_ml_io_float32_to_uint8(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_uint8_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
uint8_t *input_buffer;
@@ -194,7 +194,7 @@ rte_ml_io_uint8_to_float32(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_int16(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -226,7 +226,7 @@ rte_ml_io_float32_to_int16(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_int16_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
int16_t *input_buffer;
@@ -249,7 +249,7 @@ rte_ml_io_int16_to_float32(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_float32_to_uint16(float scale, uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -281,7 +281,7 @@ rte_ml_io_float32_to_uint16(float scale, uint64_t nb_elements, void *input, void
return 0;
}
-__rte_weak int
+int
rte_ml_io_uint16_to_float32(float scale, uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
@@ -429,7 +429,7 @@ __float32_to_float16_scalar_rtn(float x)
return u16;
}
-__rte_weak int
+int
rte_ml_io_float32_to_float16(uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -513,7 +513,7 @@ __float16_to_float32_scalar_rtx(uint16_t f16)
return f32.f;
}
-__rte_weak int
+int
rte_ml_io_float16_to_float32(uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
@@ -614,7 +614,7 @@ __float32_to_bfloat16_scalar_rtn(float x)
return u16;
}
-__rte_weak int
+int
rte_ml_io_float32_to_bfloat16(uint64_t nb_elements, void *input, void *output)
{
float *input_buffer;
@@ -696,7 +696,7 @@ __bfloat16_to_float32_scalar_rtx(uint16_t f16)
return f32.f;
}
-__rte_weak int
+int
rte_ml_io_bfloat16_to_float32(uint64_t nb_elements, void *input, void *output)
{
uint16_t *input_buffer;
--
2.17.1
^ permalink raw reply [flat|nested] 5+ messages in thread
* RE: [EXT] Re: [PATCH] mldev: remove weak symbols use in type conversions
2023-03-13 10:52 ` David Marchand
@ 2023-03-13 11:37 ` Srikanth Yalavarthi
0 siblings, 0 replies; 5+ messages in thread
From: Srikanth Yalavarthi @ 2023-03-13 11:37 UTC (permalink / raw)
To: David Marchand
Cc: dev, Shivah Shankar Shankar Narayan Rao, Srikanth Yalavarthi
> -----Original Message-----
> From: David Marchand <david.marchand@redhat.com>
> Sent: 13 March 2023 16:22
> To: Srikanth Yalavarthi <syalavarthi@marvell.com>
> Cc: dev@dpdk.org; Shivah Shankar Shankar Narayan Rao
> <sshankarnara@marvell.com>
> Subject: [EXT] Re: [PATCH] mldev: remove weak symbols use in type
> conversions
>
> External Email
>
> ----------------------------------------------------------------------
> Hello,
>
> On Sat, Mar 11, 2023 at 3:51 PM Srikanth Yalavarthi
> <syalavarthi@marvell.com> wrote:
> >
> > Drop use of weak symbols to select the type conversion functions.
> > Select NEON implementation only on Aarch64 builds. Enable windows
> > build.
> >
> > Fixes: 3c54f91ad8c9 ("mldev: disable build on Windows")
>
> Strictly speaking we are not fixing this change above, but rather
> Fixes: 9637de38a2e3 ("mldev: add scalar type conversion")
>
Updated in v2.
>
> >
> > Signed-off-by: Srikanth Yalavarthi <syalavarthi@marvell.com>
>
> The change looks good to me and the UNH Windows build target seems to
> agree :-).
Ack.
> Reviewed-by: David Marchand <david.marchand@redhat.com>
>
> Thanks Srikanth.
>
>
> --
> David Marchand
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2 1/1] mldev: remove weak symbols use in type conversions
2023-03-13 11:36 ` [PATCH v2 1/1] " Srikanth Yalavarthi
@ 2023-03-15 9:49 ` David Marchand
0 siblings, 0 replies; 5+ messages in thread
From: David Marchand @ 2023-03-15 9:49 UTC (permalink / raw)
To: Srikanth Yalavarthi; +Cc: dev, sshankarnara, aprabhu, ptakkar, pshukla
On Mon, Mar 13, 2023 at 12:37 PM Srikanth Yalavarthi
<syalavarthi@marvell.com> wrote:
>
> Drop use of weak symbols to select the type conversion
> functions. Select NEON implementation only on Aarch64
> builds. Enable windows build.
>
> Fixes: 9637de38a2e3 ("mldev: add scalar type conversion")
>
> Signed-off-by: Srikanth Yalavarthi <syalavarthi@marvell.com>
> Reviewed-by: David Marchand <david.marchand@redhat.com>
Applied, thanks.
--
David Marchand
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2023-03-15 9:50 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-03-11 14:50 [PATCH] mldev: remove weak symbols use in type conversions Srikanth Yalavarthi
2023-03-13 10:52 ` David Marchand
2023-03-13 11:37 ` [EXT] " Srikanth Yalavarthi
2023-03-13 11:36 ` [PATCH v2 1/1] " Srikanth Yalavarthi
2023-03-15 9:49 ` David Marchand
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).