From: Vlastimil Kosar
This patch adds architecture specific prefetch operations
for ARM architecture. It utilizes the pld instruction that
starts filling the appropriate cache line without blocking.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm
From: Vlastimil Kosar
This patch adds spinlock operations for ARM architecture.
We do not support HTM in spinlocks on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_spinlock.h | 114 +
1 file changed, 114
: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_memcpy.h | 270 +
1 file changed, 270 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_memcpy.h
diff --git a/lib/librte_eal/common/include/arch/arm
The GCC can be configured to avoid using NEON extensions.
For that purpose, we provide just the memcpy implementation
of the rte_memcpy.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
From: Vlastimil Kosar
This implementation is based on IBM POWER version of
rte_cpuflags. We use software emulation of HW capability
registers, because those are usually not directly accessible
from userspace on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v1 ->
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
Signed-off-by: David Hunt
---
lib/librte_eal/common/include/arch/arm/rte_cpuflags.h | 15 +++
1 file changed, 15 insertions
Just a copy from PPC.
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_rwlock.h | 40 ++
1 file changed, 40 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_rwlock.h
diff --git a/lib/librte_eal/common/include/arch/arm
There several issues with alignment when compiling for ARMv7.
They are not considered to be fatal (ARMv7 supports unaligned
access of 32b words), so we just leave them as warnings. They
should be solved later, however.
Signed-off-by: Jan Viktorin
Signed-off-by: Vlastimil Kosar
---
mk/toolchain
Signed-off-by: Jan Viktorin
---
MAINTAINERS | 4
1 file changed, 4 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 080a8e8..a8933eb 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -124,6 +124,10 @@ IBM POWER
M: Chao Zhu
F: lib/librte_eal/common/include/arch/ppc_64/
+ARM v7
+M
intrinsic functions
_mm_set_epi32(),
_mm_loadu_si128() and _mm_load_si128() for easy portability of existing
applications.
LPM builds now when on ARM.
FIXME: to be reworked
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
config/defconfig_arm-armv7-a-linuxapp-gcc | 1 -
lib
The main goal of this check is to avoid passing the -msse4.1
option to the GCC that does not support it (like arm toolchains).
Anyway, the ACL library does not compile on ARM.
Signed-off-by: Jan Viktorin
---
lib/librte_acl/Makefile | 4
1 file changed, 4 insertions(+)
diff --git a/lib
Hi Konstantin,
On Tue, 27 Oct 2015 15:31:44 +
"Ananyev, Konstantin" wrote:
> Hi Jan,
>
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Jan Viktorin
> > Sent: Monday, October 26, 2015 4:38 PM
> > To: Thoma
Jan
On Tue, 27 Oct 2015 15:55:48 +
"Ananyev, Konstantin" wrote:
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Jan Viktorin
> > Sent: Monday, October 26, 2015 4:38 PM
> > To: Thomas Monjalon; Hunt, David; dev
On Mon, 26 Oct 2015 17:22:01 +0100
Jan Viktorin wrote:
> On Fri, 23 Oct 2015 15:17:12 +0100
> David Hunt wrote:
>
> >
> > +# ARMv8 CPU flags
> > +ifneq ($(filter $(AUTO_CPUFLAGS),__aarch64__),)
I do not believe that this works. The function filter accepts
argu
d84581e3681de076819d202b1f09f2751d28d5be:
acl: handle when SSE 4.1 is unsupported (2015-10-27 20:03:24 +0100)
---
Jan Viktorin (8):
eal/arm: implement rdtsc by PMU or clock_gettime
eal/arm: use vector memcpy only when NEON is enabled
eal/arm: detect arm architecture in cpu flags
eal/arm: rwlock support for ARM
gcc
From: Vlastimil Kosar
Make DPDK run on ARMv7-A architecture. This patch assumes
ARM Cortex-A9. However, it is known to be working on Cortex-A7
and Cortex-A15.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v1 -> v2:
* the -mtune parameter of GCC is configurable now
* the -m
From: Vlastimil Kosar
This patch adds architecture specific atomic operation file
for ARM architecture. It utilizes compiler intrinsics only.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v1 -> v2:
* improve rte_wmb()
* use __atomic_* or __sync_*? (may affect the required
From: Vlastimil Kosar
This patch adds architecture specific byte order operations
for ARM. The architecture supports both big and little endian.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_byteorder.h| 148 +
1
Enable to choose a preferred way to read timer based on the
configuration entry CONFIG_RTE_ARM_EAL_RDTSC_USE_PMU.
It requires a kernel module that is not included to work.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
From: Vlastimil Kosar
ARM architecture doesn't have a suitable source of CPU cycles. This
patch uses clock_gettime instead. The implementation should be improved
in the future.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_cyc
From: Vlastimil Kosar
This patch adds architecture specific prefetch operations
for ARM architecture. It utilizes the pld instruction that
starts filling the appropriate cache line without blocking.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm
From: Vlastimil Kosar
This patch adds spinlock operations for ARM architecture.
We do not support HTM in spinlocks on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_spinlock.h | 114 +
1 file changed, 114
: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_memcpy.h | 270 +
1 file changed, 270 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_memcpy.h
diff --git a/lib/librte_eal/common/include/arch/arm
The GCC can be configured to avoid using NEON extensions.
For that purpose, we provide just the memcpy implementation
of the rte_memcpy.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
From: Vlastimil Kosar
This implementation is based on IBM POWER version of
rte_cpuflags. We use software emulation of HW capability
registers, because those are usually not directly accessible
from userspace on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
app/test
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
Signed-off-by: David Hunt
---
v2 -> v3: fixed forgotten include of string.h
---
lib/librte_eal/common/include/arch/arm/rte_cpuflags.h |
Just a copy from PPC.
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_rwlock.h | 40 ++
1 file changed, 40 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_rwlock.h
diff --git a/lib/librte_eal/common/include/arch/arm
There several issues with alignment when compiling for ARMv7.
They are not considered to be fatal (ARMv7 supports unaligned
access of 32b words), so we just leave them as warnings. They
should be solved later, however.
Signed-off-by: Jan Viktorin
Signed-off-by: Vlastimil Kosar
---
mk/toolchain
Signed-off-by: Jan Viktorin
---
MAINTAINERS | 4
1 file changed, 4 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 080a8e8..a8933eb 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -124,6 +124,10 @@ IBM POWER
M: Chao Zhu
F: lib/librte_eal/common/include/arch/ppc_64/
+ARM v7
+M
From: Vlastimil Kosar
LPM function rte_lpm_lookupx4() uses i686/x86_64 SIMD intrinsics. Therefore,
the function is reimplemented using non-vector operations for non-x86
architectures.
LPM now builds for ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v2 -> v3: as S
This patch does not map x86 SIMD operations to the ARM ones.
It just fills the necessary gap between the platforms to enable
compilation of libraries LPM (includes rte_vect.h, lpm_test needs
those SIMD functions) and ACL (includes rte_vect.h).
Signed-off-by: Jan Viktorin
---
lib/librte_eal
The main goal of this check is to avoid passing the -msse4.1
option to the GCC that does not support it (like arm toolchains).
The ACL now builds for ARM.
Signed-off-by: Jan Viktorin
---
v2 -> v3: handle missing SSE as suggested by K. Ananyev
---
lib/librte_acl/Makefile | 7 ++-
On Wed, 28 Oct 2015 11:09:21 +0100
David Marchand wrote:
> Hello Jan,
>
> On Tue, Oct 27, 2015 at 8:13 PM, Jan Viktorin
> wrote:
>
> >
> > diff --git a/config/defconfig_arm-armv7-a-linuxapp-gcc
> > b/config/defconfig_arm-armv7-a-linuxapp-gcc
> > new file
-Status: ERROR
Patchwork: http://www.dpdk.org/dev/patchwork/patch/8106/
DPDK git baseline: affc455438f4cbd3b14e2d9a24fbc154e22d68d3
Patchwork ID: 8090-8106
http://www.dpdk.org/dev/patchwork/patch/8106/
Submitter: Jan Viktorin
Date: Tue, 27 Oct 2015 20:13:49 +0100
Compilation:
OS: fedora
Nic
On Wed, 28 Oct 2015 11:17:37 +
"Ananyev, Konstantin" wrote:
> Hi Jan,
>
> > -Original Message-
> > From: Jan Viktorin [mailto:viktorin at rehivetech.com]
> > Sent: Wednesday, October 28, 2015 11:00 AM
> > To: Ananyev, Konstantin
> > C
On Wed, 28 Oct 2015 14:34:40 +0100
David Marchand wrote:
> On Mon, Oct 26, 2015 at 5:37 PM, Jan Viktorin
> wrote:
>
> > From: Vlastimil Kosar
> >
> > Make DPDK run on ARMv7-A architecture. This patch assumes
> > ARM Cortex-A9. However, it is known to be workin
On Wed, 28 Oct 2015 14:39:27 +0100
David Marchand wrote:
> On Mon, Oct 26, 2015 at 5:37 PM, Jan Viktorin
> wrote:
>
> >
> > diff --git a/config/defconfig_arm-armv7-a-linuxapp-gcc
> > b/config/defconfig_arm-armv7-a-linuxapp-gcc
> > new file mode 100644
> &
On Wed, 28 Oct 2015 13:16:24 +0100
David Marchand wrote:
> On Tue, Oct 27, 2015 at 8:13 PM, Jan Viktorin
> wrote:
>
> > There several issues with alignment when compiling for ARMv7.
> > They are not considered to be fatal (ARMv7 supports unaligned
> > access of 3
On Wed, 28 Oct 2015 15:54:47 +0100
David Marchand wrote:
> Hello Jan,
>
> On Tue, Oct 27, 2015 at 8:13 PM, Jan Viktorin
> wrote:
>
> > Hello DPDK community,
> >
> > this is the third attempt to post support for ARMv7 into the DPDK.
> > There are chang
From: Vlastimil Kosar
This patch adds architecture specific atomic operation file
for ARM architecture. It utilizes compiler intrinsics only.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v1 -> v2:
* improve rte_wmb()
* use __atomic_* or __sync_*? (may affect the required
m/RehiveTech/dpdk.git arm-support-v4
since commit 82fb702077f67585d64a07de0080e5cb6a924a72:
ixgbe: support new flow director modes for X550 (2015-10-29 00:06:01 +0100)
up to 437c85fd6d9c5f3bdd2411fb9ddf703dc4cba5a5:
maintainers: claim responsibility for ARMv7 (2015-10-29 13:33:49 +0100)
---
J
From: Vlastimil Kosar
This patch adds architecture specific byte order operations
for ARM. The architecture supports both big and little endian.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4: fix passing params to asm volatile for checkpatch
---
.../common/include/arch
From: Vlastimil Kosar
ARM architecture doesn't have a suitable source of CPU cycles. This
patch uses clock_gettime instead. The implementation should be improved
in the future.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_cyc
Enable to choose a preferred way to read timer based on the
configuration entry CONFIG_RTE_ARM_EAL_RDTSC_USE_PMU.
It requires a kernel module that is not included to work.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
From: Vlastimil Kosar
This patch adds architecture specific prefetch operations
for ARM architecture. It utilizes the pld instruction that
starts filling the appropriate cache line without blocking.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4:
* checkpatch does not like
The GCC can be configured to avoid using NEON extensions.
For that purpose, we provide just the memcpy implementation
of the rte_memcpy.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
From: Vlastimil Kosar
This patch adds spinlock operations for ARM architecture.
We do not support HTM in spinlocks on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_spinlock.h | 114 +
1 file changed, 114
: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4:
* fix whitespace issues reported by checkpatch
* fix passing params to asm volatile for checkpatch
---
.../common/include/arch/arm/rte_memcpy.h | 279 +
1 file changed, 279 insertions(+)
create mode 100644 lib
From: Vlastimil Kosar
This implementation is based on IBM POWER version of
rte_cpuflags. We use software emulation of HW capability
registers, because those are usually not directly accessible
from userspace on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
app/test
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
Signed-off-by: David Hunt
---
v2 -> v3: fixed forgotten include of string.h
v4: checkpatch reports few characters over 80 for checking aarc
Just a copy from PPC.
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_rwlock.h | 40 ++
1 file changed, 40 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_rwlock.h
diff --git a/lib/librte_eal/common/include/arch/arm
This patch does not map x86 SIMD operations to the ARM ones.
It just fills the necessary gap between the platforms to enable
compilation of libraries LPM (includes rte_vect.h, lpm_test needs
those SIMD functions) and ACL (includes rte_vect.h).
Signed-off-by: Jan Viktorin
---
v4: checkpatch
There several issues with alignment when compiling for ARMv7.
They are not considered to be fatal (ARMv7 supports unaligned
access of 32b words), so we just leave them as warnings. They
should be solved later, however.
Signed-off-by: Jan Viktorin
Signed-off-by: Vlastimil Kosar
---
v4: restrict
From: Vlastimil Kosar
Make DPDK run on ARMv7-A architecture. This patch assumes
ARM Cortex-A9. However, it is known to be working on Cortex-A7
and Cortex-A15.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v2:
* the -mtune parameter of GCC is configurable now
* the -mfpu=neon
Signed-off-by: Jan Viktorin
---
MAINTAINERS | 4
1 file changed, 4 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 080a8e8..a8933eb 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -124,6 +124,10 @@ IBM POWER
M: Chao Zhu
F: lib/librte_eal/common/include/arch/ppc_64/
+ARM v7
+M
On Thu, 29 Oct 2015 15:02:03 +0100
Thomas Monjalon wrote:
> 2015-10-28 18:58, David Marchand:
> > > > - since you introduce a new architecture, do you intend to run daily
> > > > build checks and send reports to the test-report mailing list ?
> > >
> > > I think, this is possible, if I automate
pedef Elf32_auxv_t Elf_auxv_t;
#endif
while leaving the above code almost untouched (just Elf32_auxv_t ->
Elf_auxv_t). This is like spagetti... :)
Regards
Jan
--
Jan Viktorin E-mail: Viktorin at RehiveTech.com
System Architect Web:www.RehiveTech.com
RehiveTech
Brno, Czech Republic
s.
> + * ARMv7 can be tested with virtual PMD drivers.
> + * ARMv8 can be tested with virtual and physicla PMD drivers.
Typo "physical"
>
>
> Resolved Issues
> diff --git a/mk/arch/arm64/rte.vars.mk b/mk/arch/arm64/rte.vars.mk
> new file mode 100644
--
Jan
otes/release_2_2.rst
> +++ b/doc/guides/rel_notes/release_2_2.rst
> @@ -23,10 +23,11 @@ New Features
>
> * **Added vhost-user multiple queue support.**
>
> -* **Introduce ARMv7 architecture**
--
Jan Viktorin E-mail: Viktorin at RehiveTech.com
System Archit
e_memcpy_64.h
> create mode 100644 lib/librte_eal/common/include/arch/arm/rte_prefetch_32.h
> create mode 100644 lib/librte_eal/common/include/arch/arm/rte_prefetch_64.h
> create mode 100644 mk/arch/arm64/rte.vars.mk
> create mode 100644 mk/machine/armv8a/rte.vars.mk
>
--
Jan ViktorinE-mail: Viktorin at RehiveTech.com
System ArchitectWeb:www.RehiveTech.com
RehiveTech
Brno, Czech Republic
eTech/dpdk.git arm-support-v5
since commit 82fb702077f67585d64a07de0080e5cb6a924a72:
ixgbe: support new flow director modes for X550 (2015-10-29 00:06:01 +0100)
up to 285d29f6226d53c8af8035ebaf4c9edf635e2c56:
maintainers: claim responsibility for ARMv7 (2015-10-30 01:13:26 +0100)
---
Jan Viktor
From: Vlastimil Kosar
This patch adds architecture specific atomic operation file
for ARM architecture. It utilizes compiler intrinsics only.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v1 -> v2:
* improve rte_wmb()
* use __atomic_* or __sync_*? (may affect the required
From: Vlastimil Kosar
This patch adds architecture specific byte order operations
for ARM. The architecture supports both big and little endian.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4: fix passing params to asm volatile for checkpatch
---
.../common/include/arch
From: Vlastimil Kosar
This patch adds spinlock operations for ARM architecture.
We do not support HTM in spinlocks on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_spinlock.h | 114 +
1 file changed, 114
The GCC can be configured to avoid using NEON extensions.
For that purpose, we provide just the memcpy implementation
of the rte_memcpy.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4:
* fix whitespace issues reported by checkpatch
* fix passing params to asm volatile for checkpatch
v5: prepare for applying ARMv8
---
.../common/include/arch/arm/rte_memcpy.h | 38 +++
.../common/include/arch/arm/rte_memcpy_32.h
From: Vlastimil Kosar
This implementation is based on IBM POWER version of
rte_cpuflags. We use software emulation of HW capability
registers, because those are usually not directly accessible
from userspace on ARM.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
app/test
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
Signed-off-by: Amruta Zende
Signed-off-by: David Hunt
---
v2 -> v3: fixed forgotten include of string.h
v4: checkpatch reports few characters over 80 for checking aarc
Just a copy from PPC.
Signed-off-by: Jan Viktorin
---
.../common/include/arch/arm/rte_rwlock.h | 40 ++
1 file changed, 40 insertions(+)
create mode 100644 lib/librte_eal/common/include/arch/arm/rte_rwlock.h
diff --git a/lib/librte_eal/common/include/arch/arm
This patch does not map x86 SIMD operations to the ARM ones.
It just fills the necessary gap between the platforms to enable
compilation of libraries LPM (includes rte_vect.h, lpm_test needs
those SIMD functions) and ACL (includes rte_vect.h).
Signed-off-by: Jan Viktorin
---
v4: checkpatch
There several issues with alignment when compiling for ARMv7.
They are not considered to be fatal (ARMv7 supports unaligned
access of 32b words), so we just leave them as warnings. They
should be solved later, however.
Signed-off-by: Jan Viktorin
Signed-off-by: Vlastimil Kosar
---
v4: restrict
From: Vlastimil Kosar
Make DPDK run on ARMv7-A architecture. This patch assumes
ARM Cortex-A9. However, it is known to be working on Cortex-A7
and Cortex-A15.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v2:
* the -mtune parameter of GCC is configurable now
* the -mfpu=neon
Signed-off-by: Jan Viktorin
---
MAINTAINERS | 4
1 file changed, 4 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 080a8e8..a8933eb 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -124,6 +124,10 @@ IBM POWER
M: Chao Zhu
F: lib/librte_eal/common/include/arch/ppc_64/
+ARM v7
+M
From: Vlastimil Kosar
ARM architecture doesn't have a suitable source of CPU cycles. This
patch uses clock_gettime instead. The implementation should be improved
in the future.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v5: prepare for applying ARMv8
---
.../c
Enable to choose a preferred way to read timer based on the
configuration entry CONFIG_RTE_ARM_EAL_RDTSC_USE_PMU.
It requires a kernel module that is not included to work.
Based on the patch by David Hunt and Armuta Zende:
lib: added support for armv7 architecture
Signed-off-by: Jan Viktorin
From: Vlastimil Kosar
This patch adds architecture specific prefetch operations
for ARM architecture. It utilizes the pld instruction that
starts filling the appropriate cache line without blocking.
Signed-off-by: Vlastimil Kosar
Signed-off-by: Jan Viktorin
---
v4:
* checkpatch does not like
Thanks for that hint. I am able to run it in qemu. I tried several
tests from the test suite and it works.
Jan
On Fri, 30 Oct 2015 08:52:49 +
"Hunt, David" wrote:
> On 30/10/2015 00:17, Jan Viktorin wrote:
> > I've failed to compile kni/igb for ARMv8. An
IMER_MULTIPLIER;
> +#else
I think, this should be CONFIG_RTE_TIMER_MULTIPLIER.
Jan
> + return tsc;
> +#endif
> +
> +}
> +
--
Jan Viktorin E-mail: Viktorin at RehiveTech.com
System Architect Web:www.RehiveTech.com
RehiveTech
Brno, Czech Republic
Much better then the previous...
Reviewed-by: Jan Viktorin
On Fri, 30 Oct 2015 11:51:24 +
David Hunt wrote:
> Signed-off-by: David Hunt
> ---
> lib/librte_eal/common/include/arch/arm/rte_cpuflags.h | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
>
I'd put here (in the commit message) the note that the armv8 includes
are in the arm directory.
Otherwise:
Reviewed-by: Jan Viktorin
On Fri, 30 Oct 2015 11:51:25 +
David Hunt wrote:
> Signed-off-by: David Hunt
> ---
> MAINTAINERS|
not CC'd there...
Jan?Viktorin
RehiveTech
Sent?from?a?mobile?device
? P?vodn? zpr?va ?
Od: Hunt, David
Odesl?no: p?tek, 30. ??jna 2015 15:07
Komu: Jan Viktorin
P?edm?t: Fwd: [PATCH v3 6/6] test: add checks for cpu flags on armv8
Jan,
I had gone to the trouble of adding a "Reviewed-by" li
>confused...?
Jan?Viktorin
RehiveTech
Sent?from?a?mobile?device
? P?vodn? zpr?va ?
Od: sys_stv at intel.com
Odesl?no: sobota, 31. ??jna 2015 6:53
Komu: test-report at dpdk.org; viktorin at rehivetech.com
P?edm?t: |ERROR| pw 8279-8293 maintainers: claim responsibility for ARMv7
Test-Label: In
/usr/bin/test [--no-huge] [--no-pci]
Regards
Jan Viktorin
[1] http://lists.busybox.net/pipermail/buildroot/2015-October/143413.html
[2] http://buildroot.uclibc.org/
--
Jan Viktorin E-mail: Viktorin at RehiveTech.com
System Architect Web:www.RehiveTech.
access to the
device memory specified in the device-tree's reg property, must give a way how
to perform DMA transfers, should be able to deliver interrupts, etc.
See the commits for more details...
Happy New Year!
Regards
Jan Viktorin
Jan Viktorin (7):
eal/common: define rte_soc_* relat
On Mon, 4 Jul 2016 10:22:08 +
"Burakov, Anatoly" wrote:
> > -Original Message-----
> > From: Jan Viktorin [mailto:viktorin at rehivetech.com]
> > Sent: Monday, June 13, 2016 2:02 PM
> > To: dev at dpdk.org
> > Cc: Jan Viktorin ; Burakov, Anatoly
On Wed, 29 Jun 2016 15:12:07 +0530
Shreyansh jain wrote:
> Hi Jan,
>
> On Friday 06 May 2016 07:18 PM, Jan Viktorin wrote:
> > Signed-off-by: Jan Viktorin
> > ---
> > lib/librte_ether/rte_ethdev.c | 127
> > +-
> &g
Hello Shreyansh,
On Thu, 16 Jun 2016 11:47:29 +
Shreyansh Jain wrote:
> Sorry, didn't notice this email earlier...
> Comments inline
>
> > -Original Message-
> > From: Jan Viktorin [mailto:viktorin at rehivetech.com]
> > Sent: Wednesday, June 15, 201
On Mon, 4 Jul 2016 19:57:18 +0530
Shreyansh jain wrote:
[...]
> >>> @@ -1431,7 +1524,7 @@ rte_eth_dev_info_get(uint8_t port_id, struct
> >>> rte_eth_dev_info *dev_info)
> >>>
> >>> RTE_FUNC_PTR_OR_RET(*dev->dev_ops->dev_infos_get);
> >>> (*dev->dev_ops->dev_infos_get)(dev, dev_info);
> >>
On Mon, 4 Jul 2016 10:22:08 +
"Burakov, Anatoly" wrote:
[...]
> There's no patch cover letter so I'll reply to the first patch. I've done
> some cursory testing with a NIC, nothing seems to be broken and the code
> looks OK to me. So, once this patchset is rebased on latest master (patche
Hello,
I've rebased the v2 of this patch set on top of the current master.
It builds well for my setup (both VFIO enabled and disabled).
Regards
Jan
v3:
* 0012: Acked-by: John McNamara
Jan Viktorin (16):
vfio: fix include of eal_private.h to be local
vfio: move VFIO-specific stu
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_vfio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_vfio.c
index f91b924..8b7d53f 100644
--- a/lib/librte_eal/linuxapp/eal
The common VFIO definitions should be separated from the PCI-specific parts.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 28
lib/librte_eal/linuxapp/eal/eal_vfio.h | 28
2 files changed, 28 insertions
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 7 ---
lib/librte_eal/linuxapp/eal/eal_vfio.h | 7 +++
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_vfio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_vfio.c
We make the iommu_types public temporarily here until the depending stuff is
refactored. The iommu_types and dma_map functions will be changed to be private
inside the eal_vfio module later.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/Makefile | 1 +
lib/librte_eal
The pci_vfio_set_iommu_type is not PCI-specific and it is a private function
of the eal_pci_vfio.c. We just rename the function and make it available even
for non-PCI devices.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 25 +
lib
The pci_vfio_has_supported_extensions is not PCI-specific and it is a private
function of the eal_pci_vfio.c. We just rename the function and make it
available even for non-PCI devices.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 36
The constants are not PCI-specific. Move them into the eal_vfio.h.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 7 ---
lib/librte_eal/linuxapp/eal/eal_vfio.h | 6 ++
2 files changed, 6 insertions(+), 7 deletions(-)
diff --git a/lib/librte_eal
The pci_vfio_get_container_fd is not PCI-specific. Move the implementation to
the eal_vfio.c as vfio_get_container_fd. No other code seems to call this
function.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 1 -
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c
.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 38 +-
lib/librte_eal/linuxapp/eal/eal_vfio.c | 43 ++
lib/librte_eal/linuxapp/eal/eal_vfio.h | 7 +
3 files changed, 51 insertions(+), 37 deletions(-)
diff
The setup logic access the global vfio_cfg variable that will be moved in the
following commits. We need to separate all accesses to this variable to a
general code.
Signed-off-by: Jan Viktorin
---
lib/librte_eal/linuxapp/eal/eal_pci_vfio.c | 85 +-
1 file changed
201 - 300 of 624 matches
Mail list logo
