Hi, Wei
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Wei Zhao
> Sent: Thursday, April 6, 2017 2:51 PM
> To: dev@dpdk.org
> Cc: Lu, Wenzhuo ; Zhao1, Wei
>
> Subject: [dpdk-dev] [PATCH v6 1/3] lib/librte_ether: add support for port
> reset
>
> Add support for
Hi Thomas,
On 4/7/2017 4:14 AM, Thomas Monjalon wrote:
Ping
2017-03-08 12:40, Thomas Monjalon:
2017-03-02 11:00, Ami Sabo:
The patchset fixes secondary process crash issue when it tries
to access virtio-user pmd (e.g. via rte_eth_rx_burst).
The crash happens because in virtio_user probing,
Hi Chao:
Would you help to check the PowerPC part. (i40e_rxtx_vec_altivec.c)
Thanks
Qi
> -Original Message-
> From: Zhang, Qi Z
> Sent: Friday, April 7, 2017 11:25 AM
> To: Wu, Jingjing ; Zhang, Helin
>
> Cc: Yigit, Ferruh ; jianbo@linaro.org;
> chao...@linux.vnet.ibm.com; dev
The mapping from hardware defined packet type to software defined
packet type is static for i40e device, the patch let each ethdev to
to have their own copy of mapping table, this give the possibility
that different ethdev can be set different PTYPE mapping rule which
is the requirement to support
Add below command line to configure ptype mapping.
ptype mapping get .
ptype mapping replace.
ptype mapping reset .
ptype mapping update .
Signed-off-by: Qi Zhang
---
v6:
- Update testpmd_func.rst.
v5:
- Rebase to dpdk-next-net
v3:
- Fix compile issue when CONFIG_RTE_LIBRTE_I40E_PMD=
The patch adds 4 APIs to support configurable
PTYPE mapping for i40e device.
rte_pmd_i40e_ptype_mapping_get.
rte_pmd_i40e_ptype_mapping_replace.
rte_pmd_i40e_ptype_mapping_reset.
rte_pmd_i40e_ptype_mapping_update.
The mapping from hardware defined packet type to software defined packet
type can be
The patch set create new APIs that help to change the mapping from hardware
defined packet type to software defined packet type for i40e NICs.
Keep these APIs private is because currently they are only meaningful for
devices that support dynamic PTYPE configuration, which may not be a general
devic
> -Original Message-
> From: Richardson, Bruce
> Sent: Thursday, April 6, 2017 10:37 PM
> To: Zhang, Helin
> Cc: Olivier Matz; dev@dpdk.org; Wu, Jingjing; Yigit, Ferruh
> Subject: Re: [PATCH] net/i40e: disable init and driver logs by default
>
> On Thu, Apr 06, 2017 at 03:31:23PM +0100,
> -Original Message-
> From: Olivier MATZ [mailto:olivier.m...@6wind.com]
> Sent: Thursday, April 6, 2017 11:17 PM
> To: Richardson, Bruce
> Cc: Zhang, Helin; dev@dpdk.org; Wu, Jingjing; Yigit, Ferruh
> Subject: Re: [PATCH] net/i40e: disable init and driver logs by default
>
> On Thu, 6
Hi, Nélio Laranjeiro
Thanks for your feedback.
I will change them in my next version of patch set.
-Wei
> -Original Message-
> From: Nélio Laranjeiro [mailto:nelio.laranje...@6wind.com]
> Sent: Monday, April 3, 2017 5:18 PM
> To: Dai, Wei
> Cc: dev@dpdk.org; thomas.monja...@6wind.com;
Move the call to stop the device inside the close routine because close
needs to stop the device if it isn't stopped.
Free the allocated queue buffers in close instead of doing so in remove.
Original code had these clean ups in remove which was causing memory
leak.
Signed-off-by: Sagar Abhang
--
Change comments for rte_flow_validate() function to indicate that flow
rule collision and resource validation is optional for PMD and therefore
the return codes may have different meanings.
Fixes: b1a4b4cbc0a8 ("ethdev: introduce generic flow API")
Signed-off-by: John Daley
---
v2: another crack
Adrien,
Here is another crack at the comments for rte_flow_validate. Does this
capture what you were explaining to me?
I'm still not crazy about multiple meanings for EEXIST or ENOMEM since
it makes them unusable by apps, but at least the comments try to explain it. In
17.08 what about having PM
Change comments for rte_flow_validate() function to indicate that flow
rule collision and resource validation is optional for PMD and therefore
the return codes may have different meanings.
Fixes: b1a4b4cbc0a8 ("ethdev: introduce generic flow API")
Signed-off-by: John Daley
---
lib/librte_ether
A new DPDK release candidate is ready for testing:
http://dpdk.org/browse/dpdk/tag/?id=v17.05-rc1
It is the first release candidate for DPDK 17.05,
which is going to be the biggest release ever.
There is already 782 patches so it is probably
big enough for a release candidate.
However it
2017-04-04 15:59, Andrew Rybchenko:
> If probe of the whitelisted PCI device fails, reset ret to zero
> to silently skip non-whitelisted PCI devices.
>
> Fixes: 10f6c93cea38 ("eal: do not panic on PCI failures")
>
> Signed-off-by: Andrew Rybchenko
Applied, thanks
2017-04-04 11:38, Aaron Conole:
> This function is now part of the public ABI, so should be
> advertised as such.
>
> Signed-off-by: Aaron Conole
Fixes: 37e97ad2c56a ("eal: do not panic when CPU is not supported")
Series applied, thanks
Ping,
any progress on this patch?
2017-03-23 19:36, John Daley:
> PMDs only consider if a flow would be accepted or not the assuming the
> device had all it's resources available to it. Since state is not
> considered, -EEXIST and -EBUSY return codes no longer make sense and
> are removed. Also c
2017-03-24 12:52, Andriy Berestovskyy:
> At the moment rte_eth_dev_configure() behaves inconsistent:
> - for normal frames: out of range max_rx_pkt_len uses a default
> - for jumbo frames: out of range max_rx_pkt_len gives an error
>
> This patch fixes this inconsistency by using a default value
> > The first line of the MAINTAINERS file are:
> > "
> > The intention of this file is to provide a set of names that we can rely
> > on for helping in patch reviews and questions.
> > "
> >
> > Unfortunately, some maintainers do not endorse their role for questions
> > asked on the mailing list.
2017-03-30 17:32, Wenzhuo Lu:
> VFd (the VF daemon) is a mechanism which can be used to configure
> features on a VF without direct access to the PF.
>
> Signed-off-by: Wenzhuo Lu
> Acked-by: John McNamara
Applied, thanks
2017-04-06 17:36, John McNamara:
> TILE-GX is no longer supported but the tap driver includes it in its
> NIC overview ini file which causes a doc warning:
>
> $ make doc-guides-html
> sphinx processing guides-html...
> Warning generate_nic_overview_table():
> Unknown feature 'TILE
Ping
2017-03-08 12:40, Thomas Monjalon:
> 2017-03-02 11:00, Ami Sabo:
> > The patchset fixes secondary process crash issue when it tries
> > to access virtio-user pmd (e.g. via rte_eth_rx_burst).
> >
> > The crash happens because in virtio_user probing,
> > eth_dev_attach_secondary is not being c
2017-03-29 17:38, Bruce Richardson:
> This patch aims to increase the amount of code coverage we get when
> doing a basic build of DPDK using the build tools and scripts. It does this
> by enabling in our makefiles and default build configuration extra drivers
> and examples which don't have extra
2017-03-27 15:35, Ferruh Yigit:
> Test build removed from default build, it needs to be added explicitly
> to the test-build script.
>
> Fixes: 7d3b1ec47fae ("test: move unit tests to separate directory")
>
> Signed-off-by: Ferruh Yigit
Bruce's version has been picked:
http://dpdk.org/p
2017-04-06 14:33, Yuanhan Liu:
> So that, as a stable maintainer while picking commits to a stable release,
> I could pay less attention to those have it and pay more attention to those
> don't have it.
>
> Signed-off-by: Yuanhan Liu
Acked-by: Thomas Monjalon
Applied, thanks
2017-04-06 20:42, Qi Zhang:
> It seems its not necessary to register an intr_handle for interrupt callback
> function. "void* cb_arg" shows enough when be used to pass the object that
> contain the information be required to handle the interrupt event( A typical
> way which is implemented by almost
2017-04-06 20:42, Qi Zhang:
> Remove the inappropriate modification on get_max_intr
> field that keep the intr_source read only.
>
> Signed-off-by: Qi Zhang
v2 was
Acked-by: Anatoly Burakov
2017-04-05 17:07, Ferruh Yigit:
> Signed-off-by: Ferruh Yigit
Applied, thanks
> > Failed to destroy tunnel filter rule if the action of the tunnel filter is
> > VF, root
> > cause is the wrong vsi used.
> >
> > Fixes: 5da4939e90fb ("net/i40e: support tunnel filter to VF")
> >
> > Signed-off-by: Beilei Xing
>
> Acked-by: Bernard Iremonger
Applied, thanks
2017-04-06 18:05, Ferruh Yigit:
> Disable for gcc < 4.7 and icc <= 14.0
>
> PMD uses some compiler builtins and new compiler options. Tested with
> gcc 4.5.1 and following were not supported:
>
> option:
> -Ofast
>
> macros:
> _Static_assert
>
> __ORDER_LITTLE_ENDIAN__
> __ORDER_BIG_ENDIAN__
>
2017-04-06 13:21, Shijith Thotton:
> Add a section in NIC drivers documentation to explain compiling and
> testing of a PMD. It also mentions about host setup, which is required
> before running testpmd.
>
> Add label "testpmd_ug" to refer user guide.
>
> Signed-off-by: Shijith Thotton
It is re
build error:
.../event/sw/sw_evdev_worker.c: In function ‘sw_event_release’:
.../event/sw/sw_evdev_worker.c:52:3: error: unknown field ‘op’ specified
in initializer
Fixed by updating struct initialization.
Fixes: 656af9180014 ("event/sw: add worker core functions")
Signed-off-by: Ferruh Yigit
A
-Ofast option supported from gcc4.6 [1], for older versions using
"-O3 -ffast-math" instead.
[1] build error:
cc1: error: invalid option argument ‘-Ofast’
Also struct initialization build error [2] fixed.
[2] build error:
.../event/octeontx/ssovf_mbox.c: In function ‘mbox_send_request’:
.../even
Disable for gcc < 4.7 and icc <= 14.0
PMD uses some compiler builtins and new compiler options. Tested with
gcc 4.5.1 and following were not supported:
option:
-Ofast
macros:
_Static_assert
__ORDER_LITTLE_ENDIAN__
__ORDER_BIG_ENDIAN__
__BYTE_ORDER__
__atomic_fetch_add
__ATOMIC_ACQUIRE
__atomic
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Zhang, Helin ; Wu,
> Jingjing
> Cc: dev@dpdk.org
> Subject: [PATCH 6/8] doc: refer PMD compile and test section from i
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Rasesh Mody ; Harish
> Patil
> Cc: dev@dpdk.org
> Subject: [PATCH 8/8] doc: refer PMD compile and test section from qe
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Jan Medala ; Jakub Palider
> ; Netanel Belgazal ; Evgeny
> Schemeilin
> Cc: dev@dpdk.org
> Subject: [PATCH 5/8] doc: r
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Alejandro Lucero
>
> Cc: dev@dpdk.org
> Subject: [PATCH 7/8] doc: refer PMD compile and test section from nfp doc
>
>
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Harish Patil ; Rasesh
> Mody
> Cc: dev@dpdk.org
> Subject: [PATCH 3/8] doc: refer PMD compile and test section from bn
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
> ; Rahul Lakkireddy
> Cc: dev@dpdk.org
> Subject: [PATCH 4/8] doc: refer PMD compile and test section from cxgbe
> doc
>
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
>
> Cc: dev@dpdk.org
> Subject: [PATCH 2/8] doc: refer PMD compile and test section from liquidio
> doc
>
> Refer the sec
> -Original Message-
> From: Shijith Thotton [mailto:shijith.thot...@caviumnetworks.com]
> Sent: Thursday, April 6, 2017 8:52 AM
> To: Mcnamara, John ; Yigit, Ferruh
>
> Cc: dev@dpdk.org
> Subject: [PATCH 1/8] doc: add doc to explain compiling and testing of PMD
>
> Add a section in NIC
TILE-GX is no longer supported but the tap driver includes it in its
NIC overview ini file which causes a doc warning:
$ make doc-guides-html
sphinx processing guides-html...
Warning generate_nic_overview_table():
Unknown feature 'TILE-Gx' in 'tap.ini'
Fixes: 04b2c3c94346 ("doc: a
2017-04-06 17:20, Bruce Richardson:
> On Thu, Apr 06, 2017 at 04:48:11PM +0200, Thomas Monjalon wrote:
> > It is a really nice Makefile recursive function :)
>
> "nice Makefile recursive function"
>
> I recognise those words individually, but having them together just
> doesn't make any sense :-)
Hi Bruce:
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Bruce Richardson
> Sent: Thursday, April 6, 2017 7:32 PM
> To: Zhang, Helin
> Cc: dev@dpdk.org; Yigit, Ferruh ; Richardson, Bruce
>
> Subject: [dpdk-dev] [PATCH v2 1/2] net/i40e: eliminate mbuf write on
On Thu, Apr 06, 2017 at 04:48:11PM +0200, Thomas Monjalon wrote:
> 2017-04-06 16:14, Olivier Matz:
> > I get the following error when linking the test application:
> > build/lib/librte_pmd_thunderx_nicvf.a(nicvf_hw.o):
> > In function `nicvf_qsize_regbit':
> > drivers/net/thunderx/base/nicvf_
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Chao Zhu
> Sent: Thursday, April 6, 2017 11:06 AM
> To: dev@dpdk.org
> Cc: Gowrishankar ; Gonzalez Monroy,
> Sergio ; david.march...@6wind.com
> Subject: [dpdk-dev] [PATCH 2/2] doc/guides: Add hugepage reserve
> i
The base code was using native GCC macros
for endian detection. Change to portable
rte_byteorder.h based endian detection.
Fixes: 262d43fe20ee ("net/thunderx/base: add HW register definitions")
Suggested-by: Thomas Monjalon
Signed-off-by: Jerin Jacob
---
drivers/net/thunderx/base/nicvf_hw_defs
> -Original Message-
> From: Nélio Laranjeiro [mailto:nelio.laranje...@6wind.com]
> Sent: Thursday, April 06, 2017 10:43 AM
<...>
> You are facing some kind of issue?
Ok, thanks for the clarification. No, I am no longer experiencing an issue...
the problem was between the keyboard and t
2017-04-06 16:18, Olivier MATZ:
> On Thu, 6 Apr 2017 09:59:47 -0400, Ed Czeck wrote:
> > build error:
> > include/rte_ring.h:459:22: error: invalid conversion from ‘void*’
> > to ‘void**’ [-fpermissive]
> > ENQUEUE_PTRS(r, &r[1], prod_head, obj_table, n, void *);
> >
> > Implicit casts of void
> > There is a compilation error in the following case:
> >
> > make config T=x86_64-native-linuxapp-gcc
> > sed -i 's,CONFIG_RTE_LIBRTE_CRYPTODEV=y,CONFIG_RTE_LIBRTE_CRYPTODEV=n,' \
> > build/.config
> > sed -i 's,CONFIG_RTE_APP_CRYPTO_PERF=y,CONFIG_RTE_APP_CRYPTO_PERF=n,' \
> > bui
This error has been observed with g++ 4.8 and g++ 5.4
Fixes: a6619414 ("ring: make struct and macros type agnostic")
On Thu, Apr 6, 2017 at 10:25 AM, Thomas Monjalon
wrote:
>
> Does this error happen only with C++?
>
> Please could you give a Fixes line?
On Thu, 6 Apr 2017 15:36:43 +0100, Bruce Richardson
wrote:
> On Thu, Apr 06, 2017 at 03:31:23PM +0100, Zhang, Helin wrote:
> >
> >
> > -Original Message-
> > From: Olivier Matz [mailto:olivier.m...@6wind.com]
> > Sent: Thursday, April 6, 2017 10:17 PM
> > To: dev@dpdk.org
> > Cc: Zhang
-Original Message-
> Date: Thu, 6 Apr 2017 16:23:55 +0200
> From: Olivier MATZ
> To: dev@dpdk.org, Jerin Jacob
> Cc: alejandro.luc...@netronome.com
> Subject: Re: [dpdk-dev] [PATCH] common: introduce an integer log2 function
> X-Mailer: Claws Mail 3.14.1 (GTK+ 2.24.31; x86_64-pc-linux-gnu
-Original Message-
> Date: Thu, 6 Apr 2017 16:13:37 +0200
> From: Olivier Matz
> To: dev@dpdk.org
> CC: jerin.ja...@caviumnetworks.com
> Subject: [PATCH] test/eventdev: fix compilation with crypto disabled
> X-Mailer: git-send-email 2.11.0
>
> There is a compilation error in the following
2017-04-06 16:14, Olivier Matz:
> I get the following error when linking the test application:
> build/lib/librte_pmd_thunderx_nicvf.a(nicvf_hw.o):
> In function `nicvf_qsize_regbit':
> drivers/net/thunderx/base/nicvf_hw.c:451: undefined reference to `log2'
> build/lib/librte_pmd_thunderx_n
On 4/6/2017 3:09 PM, Thomas Monjalon wrote:
> 2017-04-06 15:05, Ferruh Yigit:
>> On 4/6/2017 2:59 PM, Thomas Monjalon wrote:
>>> 2017-04-06 14:46, Ferruh Yigit:
--- a/mk/toolchain/gcc/rte.toolchain-compat.mk
+++ b/mk/toolchain/gcc/rte.toolchain-compat.mk
@@ -89,4 +89,8 @@ else
On Wed, Apr 05, 2017 at 08:23:35PM +, Legacy, Allain wrote:
> Hi,
> None of the comments in the rte_flow.h file (or the programmers guide)
> specify what endianness should be applied to spec/mask fields. Based
> on the testing I have done so far using a CX4 device (mlx5 driver)
> fields like V
Hi Helin,
Thanks for acknowledging.
This mail to explain how important it is to configure his mailer.
You need to quote the original mail so we can quickly find what
you have replied.
It is also a problem for patchwork as it creates a new entry:
http://dpdk.org/dev/patchwork/patch/23300/
On Thu, Apr 06, 2017 at 03:31:23PM +0100, Zhang, Helin wrote:
>
>
> -Original Message-
> From: Olivier Matz [mailto:olivier.m...@6wind.com]
> Sent: Thursday, April 6, 2017 10:17 PM
> To: dev@dpdk.org
> Cc: Zhang, Helin ; Wu, Jingjing
> ; Richardson, Bruce ;
> Yigit, Ferruh
> Subject:
2017-04-06 16:12, Olivier Matz:
> On my system, the version of the compiler is not properly retrieved,
> resulting in strange logs when building the dpdk:
> /bin/sh: line 0: test: too many arguments
>
> This happens when mk/toolchain/clang/rte.toolchain-compat.mk is included
> from a directory t
-Original Message-
From: Olivier Matz [mailto:olivier.m...@6wind.com]
Sent: Thursday, April 6, 2017 10:17 PM
To: dev@dpdk.org
Cc: Zhang, Helin ; Wu, Jingjing ;
Richardson, Bruce ; Yigit, Ferruh
Subject: [PATCH] net/i40e: disable init and driver logs by default
Since "net/i40e: use dy
2017-04-06 14:57, Thomas Monjalon:
> The file examples.dox was not re-generated when a file
> is added or removed from examples/.
> It is now removed on clean operation.
> The ordering of operations (clean before generation) is also
> better defined.
>
> Signed-off-by: Thomas Monjalon
Applied qu
2017-04-06 16:18, Pascal Mazon:
> This macro was defined twice.
> On machines with kernel < 3.8, IFF_MULTI_QUEUE didn't exist, and thus
> both definitions used different values.
>
> Fixes: cf5643661161 ("net/tap: move private elements to external header")
>
> Signed-off-by: Pascal Mazon
Applied
2017-04-06 16:18, Olivier MATZ:
> On Thu, 6 Apr 2017 09:59:47 -0400, Ed Czeck wrote:
> > build error:
> > include/rte_ring.h:459:22: error: invalid conversion from ‘void*’
> > to ‘void**’ [-fpermissive]
> > ENQUEUE_PTRS(r, &r[1], prod_head, obj_table, n, void *);
> >
> > Implicit casts of void
I forgot to CC Alejandro for nfp.
Also CC Jerin: I think the same could be done in thunderx, but
I did not want to do the patch since it changes the base/ driver.
On Thu, 6 Apr 2017 16:15:36 +0200, Olivier Matz wrote:
> At some places, the log2() function is used despite this function
> works
On Thu, 6 Apr 2017 09:59:47 -0400, Ed Czeck wrote:
> build error:
> include/rte_ring.h:459:22: error: invalid conversion from ‘void*’
> to ‘void**’ [-fpermissive]
> ENQUEUE_PTRS(r, &r[1], prod_head, obj_table, n, void *);
>
> Implicit casts of void* to void** are considered warnings in some
>
This macro was defined twice.
On machines with kernel < 3.8, IFF_MULTI_QUEUE didn't exist, and thus
both definitions used different values.
Fixes: cf5643661161 ("net/tap: move private elements to external header")
Signed-off-by: Pascal Mazon
---
drivers/net/tap/rte_eth_tap.c | 6 --
drivers
Since "net/i40e: use dynamic log type for control logs", the
i40e driver is more verbose by default, which could result in
testpmd being flooded by log messages in some conditions:
Checking link statuses...
i40e_dev_handle_aq_msg(): Request 2561 is not supported yet
i40e_dev_handle_aq_msg():
At some places, the log2() function is used despite this function
works on float. This introduces a dependency to the math lib but
most of the time it is not required because we want an integer log2.
Add a new helper to do this job and fix 2 drivers.
Signed-off-by: Olivier Matz
---
drivers/net/
I get the following error when linking the test application:
build/lib/librte_pmd_thunderx_nicvf.a(nicvf_hw.o):
In function `nicvf_qsize_regbit':
drivers/net/thunderx/base/nicvf_hw.c:451: undefined reference to `log2'
build/lib/librte_pmd_thunderx_nicvf.a(nicvf_hw.o):
In function `nicvf_r
There is a compilation error in the following case:
make config T=x86_64-native-linuxapp-gcc
sed -i 's,CONFIG_RTE_LIBRTE_CRYPTODEV=y,CONFIG_RTE_LIBRTE_CRYPTODEV=n,' \
build/.config
sed -i 's,CONFIG_RTE_APP_CRYPTO_PERF=y,CONFIG_RTE_APP_CRYPTO_PERF=n,' \
build/.config
make -j4
make
On my system, the version of the compiler is not properly retrieved,
resulting in strange logs when building the dpdk:
/bin/sh: line 0: test: too many arguments
This happens when mk/toolchain/clang/rte.toolchain-compat.mk is included
from a directory that use gcc to build (ex: kernel modules). I
2017-04-06 15:05, Ferruh Yigit:
> On 4/6/2017 2:59 PM, Thomas Monjalon wrote:
> > 2017-04-06 14:46, Ferruh Yigit:
> >> --- a/mk/toolchain/gcc/rte.toolchain-compat.mk
> >> +++ b/mk/toolchain/gcc/rte.toolchain-compat.mk
> >> @@ -89,4 +89,8 @@ else
> >>ifeq ($(shell test $(GCC_VERSION) -lt 42 && e
2017-04-06 12:00, Pascal Mazon:
> Fixes: cf5643661161 ("net/tap: move private elements to external header")
>
> Signed-off-by: Pascal Mazon
Please, could you describe the bug briefly?
I know it is a build error. In this case, a log is appreciated.
Why have I not seen the build issue? Is it speci
On 4/6/2017 2:59 PM, Thomas Monjalon wrote:
> 2017-04-06 14:46, Ferruh Yigit:
>> --- a/mk/toolchain/gcc/rte.toolchain-compat.mk
>> +++ b/mk/toolchain/gcc/rte.toolchain-compat.mk
>> @@ -89,4 +89,8 @@ else
>> ifeq ($(shell test $(GCC_VERSION) -lt 42 && echo 1), 1)
>> MACHINE_CFLAGS
-Original Message-
> Date: Thu, 06 Apr 2017 15:52:40 +0200
> From: Thomas Monjalon
> To: Ferruh Yigit , Jerin Jacob
>
> Cc: Bruce Richardson , Harry van Haaren
> , dev@dpdk.org
> Subject: Re: [PATCH 1/3] net/thunderx: disable pmd for gcc < 4.7
> User-Agent: KMail/4.14.10 (Linux/4.5.4-1-
build error:
include/rte_ring.h:459:22: error: invalid conversion from ‘void*’
to ‘void**’ [-fpermissive]
ENQUEUE_PTRS(r, &r[1], prod_head, obj_table, n, void *);
Implicit casts of void* to void** are considered warnings in some
compilers. E.g. g++ version 5.8. Cast directly to object types
S
2017-04-06 14:46, Ferruh Yigit:
> --- a/mk/toolchain/gcc/rte.toolchain-compat.mk
> +++ b/mk/toolchain/gcc/rte.toolchain-compat.mk
> @@ -89,4 +89,8 @@ else
> ifeq ($(shell test $(GCC_VERSION) -lt 42 && echo 1), 1)
> MACHINE_CFLAGS := $(filter-out -march% -mtune%
> -msse%,$(MACHI
2017-04-06 13:14, Ferruh Yigit:
> PMD uses some compiler builtins and new compiler options. Tested with
> gcc 4.5.1 and following were not supported:
>
> option:
> -Ofast
>
> macros:
> _Static_assert
>
> __ORDER_LITTLE_ENDIAN__
> __ORDER_BIG_ENDIAN__
> __BYTE_ORDER__
These ones should not be us
build error:
.../event/sw/sw_evdev_worker.c: In function ‘sw_event_release’:
.../event/sw/sw_evdev_worker.c:52:3: error: unknown field ‘op’ specified
in initializer
Fixed by updating struct initialization.
Fixes: 656af9180014 ("event/sw: add worker core functions")
Signed-off-by: Ferruh Yigit
A
Disable for gcc < 4.7 and icc <= 14.0
PMD uses some compiler builtins and new compiler options. Tested with
gcc 4.5.1 and following were not supported:
option:
-Ofast
macros:
_Static_assert
__ORDER_LITTLE_ENDIAN__
__ORDER_BIG_ENDIAN__
__BYTE_ORDER__
__atomic_fetch_add
__ATOMIC_ACQUIRE
__atomic
-Ofast option supported from gcc4.6 [1], for older versions using
"-O3 -ffast-math" instead.
[1] build error:
cc1: error: invalid option argument ‘-Ofast’
Also struct initialization build error [2] fixed.
[2] build error:
.../event/octeontx/ssovf_mbox.c: In function ‘mbox_send_request’:
.../even
On 4/6/2017 1:14 PM, Ferruh Yigit wrote:
> PMD uses some compiler builtins and new compiler options. Tested with
> gcc 4.5.1 and following were not supported:
>
> option:
> -Ofast
>
> macros:
> _Static_assert
>
> __ORDER_LITTLE_ENDIAN__
> __ORDER_BIG_ENDIAN__
> __BYTE_ORDER__
>
> __atomic_fetch
Hi Chao,
You mentioned that 'mmap will not respect the requested address hint',
how does the proposed change solves that?
Is it that hugepages map to a specific VA region, and without
MAP_HUGETLB you may get address from wrong region?
If mmap were to respect the hinted address, we could do
The file examples.dox was not re-generated when a file
is added or removed from examples/.
It is now removed on clean operation.
The ordering of operations (clean before generation) is also
better defined.
Signed-off-by: Thomas Monjalon
---
mk/rte.sdkdoc.mk | 7 ---
1 file changed, 4 inserti
Sorry, wrong patch
---
devtools/test-build.sh | 2 ++
1 file changed, 2 insertions(+)
diff --git a/devtools/test-build.sh b/devtools/test-build.sh
index 5fed4c1..d7b5970 100755
--- a/devtools/test-build.sh
+++ b/devtools/test-build.sh
@@ -230,6 +230,8 @@ for conf in $configs ; do
make -j$J EXTRA_CFLAGS="$ma
-Original Message-
> Date: Thu, 6 Apr 2017 13:14:26 +0100
> From: Ferruh Yigit
> To: Jerin Jacob , Bruce Richardson
> , Harry van Haaren
> , Thomas Monjalon
> Cc: dev@dpdk.org, Ferruh Yigit
> Subject: [PATCH 1/3] net/thunderx: disable pmd for gcc < 4.7
> X-Mailer: git-send-email 2.8.4
-Original Message-
> Date: Thu, 6 Apr 2017 13:14:27 +0100
> From: Ferruh Yigit
> To: Jerin Jacob , Bruce Richardson
> , Harry van Haaren
> , Thomas Monjalon
> Cc: dev@dpdk.org, Ferruh Yigit
> Subject: [PATCH 2/3] event/octeontx: fix compilation for gcc < 4.6
> X-Mailer: git-send-email
> From: Yigit, Ferruh
> Sent: Thursday, April 6, 2017 1:14 PM
> To: Jerin Jacob ; Richardson, Bruce
> ; Van Haaren, Harry ;
> Thomas Monjalon
>
> Cc: dev@dpdk.org; Yigit, Ferruh
> Subject: [PATCH 3/3] event/sw: fix build error for gcc 4.5.1
>
> build error:
> .../event/sw/sw_evdev_worker.c: In
build error:
.../event/sw/sw_evdev_worker.c: In function ‘sw_event_release’:
.../event/sw/sw_evdev_worker.c:52:3: error: unknown field ‘op’ specified
in initializer
Fixed by updating struct initialization.
Fixes: 656af9180014 ("event/sw: add worker core functions")
Signed-off-by: Ferruh Yigit
-
-Ofast option supported from gcc4.6 [1], for older versions using
"-O3 -ffast-math" instead.
[1] build error:
cc1: error: invalid option argument ‘-Ofast’
Also struct initialization build error [2] fixed.
[2] build error:
.../event/octeontx/ssovf_mbox.c: In function ‘mbox_send_request’:
.../even
PMD uses some compiler builtins and new compiler options. Tested with
gcc 4.5.1 and following were not supported:
option:
-Ofast
macros:
_Static_assert
__ORDER_LITTLE_ENDIAN__
__ORDER_BIG_ENDIAN__
__BYTE_ORDER__
__atomic_fetch_add
__ATOMIC_ACQUIRE
__atomic_load_n
__ATOMIC_RELAXED
__atomic_store
Some applications like OVS knows nothing about the
device type and wants to use same API to work with
all of them. But bond_pmd, unlike other pmds, requires
additional step (removing of all the slaves) before
closing the device.
In fact that bond_pmd automatically adds all the
devices from kvargs
Hi Umesh,
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Umesh Kartha
> Sent: Wednesday, March 22, 2017 10:17 AM
> To: dev@dpdk.org
> Cc: Jerin Jacob ; Balasubramanian
> Manoharan ; Ram Kumar
> ; Murthy Nidadavolu
> ; Doherty, Declan
> ; De Lara Guarch, Pablo
>
Having packets received without any offload flags given in the mbuf is not
very useful, and performance tests with testpmd indicates little to no
benefit is got with the current code by turning off the flags. This makes
the build-time option pointless, so we can remove it.
Signed-off-by: Bruce Ric
With the mbuf rework, we now have 8 contiguous bytes to be rearmed in the
mbuf just before the 8-bytes of olflags. If we don't do the rearm write
inside the descriptor ring replenishment function, and delay it to
receiving the packet, we can do a single 16B write inside the RX function
to set both
this set is based upon Olivier's mbuf rework patchset, and makes some
improvement to the i40e driver taking account of the rework. It also
removes a build-time option that seems unnecessary.
V2: Eliminate one checkpatch warning, and remove #ifdef block from
new altivec driver code.
Bruce Rich
1 - 100 of 141 matches
Mail list logo
