On 10/21/24 10:47 AM, Konstantin Ananyev wrote:
NOTE UPFRONT: this version is still not ready for merging.
Missing items:
- ARM/PPC tests passing
- PG update
SORING tests pass on a P9 PowerNV system. (Results attached for review.)
Original ring_stress_autotest still fails as documented in Bug
behavior without the patch).
Bugzilla ID: 1221
Suggested-by: Thomas Monjalon
Signed-off-by: David Christensen
Reviewed-by: David Marchand
---
v7:
* Fix space/indent errors
v6:
* Add Windows OS, fix copyright date, use EAL_LOG(), fix "stdio.h" include
v5:
* Rebase patch to current releas
behavior without the patch).
Bugzilla ID: 1221
Suggested-by: Thomas Monjalon
Signed-off-by: David Christensen
Reviewed-by: David Marchand
---
v6:
* Add Windows OS, fix copyright date, use EAL_LOG(), fix "stdio.h" include
v5:
* Rebase patch to current release, fix new checkpatch w
On 5/29/24 11:32 AM, David Christensen wrote:
IBM POWER systems support more than one type of memory management unit
(MMU). The Power ISA 3.0 specification, which applies to P9 and later
CPUs, defined a new Radix MMU which, among other things, allows an
anonymous memory page mapping to be
Modify email address to route email traffic through different
internal email servers.
Signed-off-by: David Christensen
---
MAINTAINERS | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index c9adff9846..ecc06c1216 100644
--- a/MAINTAINERS
+++ b
From: David Christensen
IBM POWER systems support more than one type of memory management unit
(MMU). The Power ISA 3.0 specification, which applies to P9 and later
CPUs, defined a new Radix MMU which, among other things, allows an
anonymous memory page mapping to be converted into a hugepage
Signed-off-by: David Christensen
---
lib/eal/arm/meson.build | 1 +
lib/eal/arm/rte_mmu.c | 11 ++
lib/eal/common/eal_private.h | 7
lib/eal/freebsd/eal.c | 7
lib/eal/linux/eal.c | 7
lib/eal/loongarch/meson.build | 1 +
lib/eal/loongarch
On 11/24/23 2:09 AM, christian.ehrha...@canonical.com wrote:
From: David Wilder
When using --no-huge option physical address are not guaranteed
to be persistent.
This change effectively makes "--no-huge" the same as
"--no-huge --iova-mode=va".
When --no-huge is used (or any other condition
address. This is a required feature in DPDK so we need
to test the MMU type when POWER systems are used and provide a more
useful error message for the user when running on an unsupported
system.
Bugzilla ID: 1221
Suggested-by: Thomas Monjalon
Signed-off-by: David Christensen
---
v3:
* Move test
On 10/23/23 4:19 PM, David Christensen wrote:
diff --git a/lib/eal/common/eal_private.h b/lib/eal/common/eal_private.h
index ebd496b537cf..7d84adb5b328 100644
--- a/lib/eal/common/eal_private.h
+++ b/lib/eal/common/eal_private.h
@@ -354,6 +354,13 @@ unsigned eal_cpu_core_id(unsigned lcore_id
address. This is a required feature in DPDK so we need
to test the MMU type when POWER systems are used and provide a more
useful error message for the user when running on an unsupported
system.
Bugzilla ID: 1221
Suggested-by: Thomas Monjalon
Signed-off-by: David Christensen
---
v2:
* Replace ifdef
On 10/17/23 5:39 AM, Thomas Monjalon wrote:
I feel this function should not be implemented in the common EAL.
What about adding a new function in lib/eal/ppc/ ?
And add the "return true" for other architectures?
Would it be more appropriate in the lib/eal/common level or
lib/eal/linux only?
address. This is a required feature in DPDK so we need
to test the MMU type when POWER systems are used and provide a more
useful error message for the user when running on an unsupported
system.
Bugzilla ID: 1221
Suggested-by: Thomas Monjalon
Signed-off-by: David Christensen
---
lib/eal/linux/eal.c
On 9/29/23 6:48 AM, Ferruh Yigit wrote:
On 6/7/2023 7:47 PM, Ferruh Yigit wrote:
On 5/16/2023 10:55 AM, Ferruh Yigit wrote:
On 5/16/2023 2:28 AM, Stephen Hemminger wrote:
On Tue, 16 May 2023 00:35:56 +0100
Ferruh Yigit wrote:
Yes only some scripts and possible applications that hotplug t
host: add VDUSE device startup")
Cc: maxime.coque...@redhat.com
Cc: sta...@dpdk.org
Signed-off-by: Bruce Richardson
---
Tested-by: David Christensen
than 11.4.
Bugzilla ID: 1251
Signed-off-by: David Christensen
Tested-by: Thinh Tran
---
v2:
- Fixed spelling error in commit message
---
config/ppc/meson.build | 8
1 file changed, 8 insertions(+)
diff --git a/config/ppc/meson.build b/config/ppc/meson.build
index 1cba44011f..160b203cb1 10
than 11.4.
Bugzilla ID: 1251
Signed-off-by: David Christensen
Tested-by: Thinh Tran
---
config/ppc/meson.build | 8
1 file changed, 8 insertions(+)
diff --git a/config/ppc/meson.build b/config/ppc/meson.build
index 1cba44011f..160b203cb1 100644
--- a/config/ppc/meson.build
+++ b/config/ppc/m
Modify LOCAL_MAC value to match new locally administered MAC
address used by TAP driver.
Bugzilla ID: 1198
Signed-off-by: David Christensen
---
examples/ipsec-secgw/test/common_defs.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/examples/ipsec-secgw/test/common_defs.sh
e set to 1, resulting in
the new address 02:64:74:61:70:.
Bugzilla ID: 1198
Signed-off-by: David Christensen
---
doc/guides/nics/tap.rst | 8
drivers/net/tap/rte_eth_tap.c| 4 ++--
examples/ipsec-secgw/test/common_defs.sh | 2 +-
3 files changed, 7 insertions(+), 7
Building DPDK with gcc 12 on a ppc64le system generates a
stringop-overflow warning. Replace the local MAC address
validation function parse_user_mac() with a call to
rte_ether_unformat_addr() instead.
Bugzilla ID: 1197
Cc: sta...@dpdk.org
Signed-off-by: David Christensen
---
v2:
* Added NULL
On 3/22/23 4:43 PM, Stephen Hemminger wrote:
static int
set_mac_type(const char *key __rte_unused,
const char *value,
@@ -2311,7 +2288,7 @@ set_mac_type(const char *key __rte_unused,
goto success;
}
- if (parse_user_mac(user_mac, value) != 6)
+ i
Building DPDK with gcc 12 on a ppc64le system generates a
stringop-overflow warning. Replace the local MAC address
validation function parse_user_mac() with a call to
rte_ether_unformat_addr() instead.
Bugzilla ID: 1197
Cc: sta...@dpdk.org
Signed-off-by: David Christensen
---
drivers/net/tap
: David Christensen
Tested-by: Thinh Tran
---
lib/acl/acl_run_altivec.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/acl/acl_run_altivec.h b/lib/acl/acl_run_altivec.h
index 4dfe7a14b4..4556e1503b 100644
--- a/lib/acl/acl_run_altivec.h
+++ b/lib/acl/acl_run_altivec.h
On 9/2/22 2:18 AM, pbhagavat...@marvell.com wrote:
From: Pavan Nikhilesh
Fix port group mask generation in altivec, vec_any_eq returns
0 or 1 while port_groupx4 expects comparison mask result.
Fixes: 2193b7467f7a ("examples/l3fwd: optimize packet processing on powerpc")
Cc: sta...@dpdk.org
On 8/29/22 3:30 AM, Thomas Monjalon wrote:
What is the conclusion on this patch?
It is good to go? Acked?
Not from me yet.
Just asked about the test environment so I can duplicate the issue. My
understanding is that Ubuntu cross-compiles for CI/CD are still working
so I'd like to underst
On 7/7/22 4:41 AM, Ali Alnubani wrote:
Meson fails to detect the dependencies that are included
in PKG_CONFIG_PATH and built for ppc64le if binaries.pkgconfig
is not set in the ppc64le cross-file for Ubuntu.
This fixes the issue by setting binaries.pkgconfig to the
binary provided by the package
On 7/18/22 1:16 PM, Thinh Tran wrote:
On 7/14/2022 1:58 AM, Ali Alnubani wrote:
-Original Message-
From: Thinh Tran
Sent: Wednesday, July 13, 2022 10:50 PM
To: Ali Alnubani ; NBU-Contact-Thomas Monjalon
(EXTERNAL)
Cc: dev@dpdk.org; David Christensen
Subject: Re: [PATCH] config
On 7/19/22 8:26 AM, Morten Brørup wrote:
This RFC proposes a set of functions optimized for non-temporal memory copy.
At this stage, I am asking for feedback on the concept.
Applications sometimes data to another memory location, which is only used
much later.
In this case, it is inefficient
On Thu, Jun 9, 2022 at 2:17 PM Stanislaw Kardach wrote:
As David noticed in [1] there is an issue with C++ compilation of the
rte_vect.h header in RISC-V. Upon closer inspection, the problem appears on
all architectures due to the type conversion rules in C++.
More precisely a union type rte_xm
+ fclose(f);
+out:
+ return (uint64_t) base;
+#else
+ return 0;
+#endif
+
}
Reviewed-by: David Christensen
2 insertions(+), 692 deletions(-)
Reviewed-by: David Christensen
On 5/2/22 7:26 AM, Duncan Bellamy wrote:
musl lacks __ppc_get_timebase() but has __builtin_ppc_get_timebase()
the __ppc_get_timebase_freq() is taken from:
https://git.alpinelinux.org/aports/commit/?id=06b03f70fb94972286c0c9f6278df89e53903833
Signed-off-by: Duncan Bellamy
---
lib/eal/ppc/i
Add support for Address Sanitizer (ASan) for PPC/POWER architecture.
Signed-off-by: David Christensen
---
v2: Removed architecture test, modified programmer's guide text
config/meson.build | 2 +-
doc/guides/prog_guide/asan.rst | 6 +++---
lib/eal/common/malloc_elem.h | 2
Add support for Address Sanitizer (ASan) for PPC/POWER architecture.
Signed-off-by: David Christensen
---
config/meson.build | 2 +-
doc/guides/prog_guide/asan.rst | 4 ++--
lib/eal/common/malloc_elem.h | 2 ++
3 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/config
Not sure if below patch also help to solve the issue you met on power
arch,
Looks like the error is same: "writing 1 byte into a region of size 0"
would you help to check?
commit 1b0f3a18145468c309bb5a8cb98a8aa29af059e7
Author: Ferruh Yigit
Date: Fri Oct 29 11:37:01 20
Enable Github Actions to cross-compile code for POWER systems.
Signed-off-by: David Christensen
Reviewed-by: David Marchand
---
v2 - Changed cross compile filename to be consistent with ARM
.ci/linux-build.sh | 6 +-
.github/workflows/build.yml
Enable Github Actions to cross-compile code for POWER systems.
Signed-off-by: David Christensen
---
.ci/linux-build.sh | 6 +-
.github/workflows/build.yml | 15 ++-
config/ppc/ppc64le-power8-linux-gcc-GHA | 11 +++
3 files changed, 30
ow;
+ }
+ if (unlikely(rxq->hw_timestamp)) {
int offset = rxq->timestamp_offset;
if (rxq->rt_timestamp) {
struct mlx5_dev_ctx_shared *sh = rxq->sh;
Reviewed-by: David Christensen
], 0LL};
cqes[1] = (vector unsigned char)
vec_sel((vector unsigned short)cqes[1],
(vector unsigned short)cqe_tmp2, cqe_sel_mask2);
Reviewed-by: David Christensen
On 10/26/21 12:08 AM, liaobiting wrote:
Hi: Please help to see this DPDK problem. And I want to know whether
i40e vf support promisc or not. Thanks a lot. From: liaobiting
Subject: Please help to see this DPDK
problem//Reply: DPDK 20.11 - i40evf: No ZjQcmQRYFpfptBannerStart
This Message Is
or IBM Power10")
Cc: thin...@linux.vnet.ibm.com
Signed-off-by: David Christensen
---
config/ppc/meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/config/ppc/meson.build b/config/ppc/meson.build
index cba8222163..1cba44011f 100644
--- a/config/ppc/meson.build
+++ b/c
dev_info->max_tx_queues = ETH_MEMIF_MAX_NUM_Q_PAIRS;
dev_info->min_rx_bufsize = 0;
Tested-by: David Christensen
On 9/22/21 2:52 PM, Dharmik Thakkar wrote:
Convert rte_atomic usages to compiler atomic built-ins
for stats sync
Signed-off-by: Dharmik Thakkar
Reviewed-by: Joyce Kong
Reviewed-by: Ruifeng Wang
---
Tested-by: David Christensen
project (2) but can be resolved
by preventing the compiler from unrolling the loop as part of the -O3
optimization.
(1) https://www.ibm.com/support/pages/advance-toolchain-linux-power
(2) https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102316
Bugzilla ID: 743
Signed-off-by: David Christensen
-
lib/eal/unix/rte_thread.c | 1 -
lib/eal/x86/rte_cycles.c | 1 -
44 files changed, 2 insertions(+), 203 deletions(-)
Reviewed-by: David Christensen
On 10/4/21 3:10 AM, Sean Morrissey wrote:
These header includes have been flagged by the iwyu_tool
and removed.
Signed-off-by: Sean Morrissey
---
lib/eal/common/eal_common_dev.c| 5 -
lib/eal/common/eal_common_devargs.c| 1 -
lib/eal/common/eal_common_errno.c | 4
On 9/30/21 2:07 AM, Min Hu (Connor) wrote:
Hi, all,
I got the answer: testpmd fwd mode should be set "io", then ping
OK.
IO fwd mode, it will not change packet MAC address info, and
ping OK
MAC fwd mode, it will change packet MAC address info, like:
"rte_ether_addr_copy(&p
On 9/29/21 12:51 AM, De Lara Guarch, Pablo wrote:
Hi
-Original Message-
From: David Christensen
Sent: Tuesday, September 28, 2021 4:53 PM
To: De Lara Guarch, Pablo ; Wang, Yipeng1
; Marohn, Byron
Cc: dev@dpdk.org; Mcnamara, John
Subject: Re: [PATCH v4] efd: change data type of
On 9/28/21 6:58 AM, Pablo de Lara wrote:
rte_efd_create() function was using uint8_t for a socket bitmask,
for one of its parameters.
This limits the maximum of NUMA sockets to be 8.
Changing to to uint64_t increases it to 64, which should be
more future-proof.
Coverity issue: 366390
Fixes: 5
On 9/18/21 12:21 AM, Peng, ZhihongX wrote:
-Original Message-
From: David Christensen
Sent: Saturday, September 18, 2021 4:51 AM
To: Peng, ZhihongX ; Burakov, Anatoly
; Ananyev, Konstantin
; step...@networkplumber.org
Cc: dev@dpdk.org; Lin, Xueqin
Subject: Re: [dpdk-dev] [PATCH
On 9/17/21 5:56 AM, Pablo de Lara wrote:
rte_efd_create() function was using uint8_t for a socket bitmask,
for one of its parameters.
This limits the maximum of NUMA sockets to be 8.
Changing to uint64_t increases it to 64, which should be
more future-proof.
Coverity issue: 366390
Fixes: 56b6
If you want to use this feature,
you need to add below compilation options when compiling code:
-Dbuildtype=debug -Db_lundef=false -Db_sanitize=address
"-Dbuildtype=debug": Display code information when coredump occurs in
the program.
"-Db_lundef=false": It is enabled by default, and needs to be d
his allows end applications to set their own
"march" value while still ensuring valid DPDK compilation.
Signed-off-by: Bruce Richardson
---
CC: Jerin Jacob
CC: Ruifeng Wang
CC: Honnappa Nagarahalli
CC: David Christensen for
Feedback requested from ARM and PPC maintainers as to this
On 9/14/21 12:10 AM, David Marchand wrote:
On Fri, Sep 10, 2021 at 6:54 PM Pablo de Lara
wrote:
rte_efd_create() function was using uint8_t for a socket bitmask,
for one of its parameters.
This limits the maximum of NUMA sockets to be 8.
Changing to to uint64_t increases it to 64, which sho
From: Zhihong Peng
AddressSanitizer (ASan) is a google memory error detect
standard tool. It could help to detect use-after-free and
{heap,stack,global}-buffer overflow bugs in C/C++ programs,
print detailed error information when error happens, large
improve debug efficiency.
By referring to i
://www.dpdk.org/blog/2021/03/26/dpdk-adopts-the-c11-memory-model/
[2] https://doc.dpdk.org/guides/rel_notes/deprecation.html
[3] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98519
Signed-off-by: David Christensen
---
lib/eal/ppc/include/rte_atomic.h | 231 +++
1 file
Piotr says lscpu isn't available on ppc FreeBSD. I think this leaves us with
two options:
1. either use sysctl with the risk that it may not return an accurate value
2. not support PPC for numa detection
What do you think?
I think point #2 is more not supporting FreeBSD on PPC for numa dete
failure. Modify the comparison to compare uint64_t values.
Fixes: fa3253c534b1 ("test/atomic: add 128-bit atomic compare exchange test")
Cc: phil.y...@arm.com
Signed-off-by: David Christensen
Tested-by: Ruifeng Wang
Reviewed-by: Ruifeng Wang
---
v2:
- Fixed commit message spel
eal/x86: ignore gcc 10 stringop-overflow warnings")
Signed-off-by: David Christensen
---
v2:
- update copyright year
- rebase for 21.11-rc0
David M,
Any chance of getting this approved in the 21.11 cycle? Patch was
originally submitted on 6/23, is only applicable to PPC architecture,
a
failure. Modify the comparison to compare uint64_t values.
Fixes: fa3253c534b1 ("test/atomic: add 128-bit atomic compare exchange test")
Cc: phil.y...@arm.com
Signed-off-by: David Christensen
Tested-by: Ruifeng Wang
Reviewed-by: Ruifeng Wang
---
v2:
- Fixed commit message spel
the vector alignment changes introduced in GCC 5. Older gcc
versions forced vector alignment to 16 bytes due to requirements for
POWER 6 and earlier CPUs, but these restrictions don't apply to CPUs
supported by DPDK.
Bugzilla ID: 739
Signed-off-by: David Christensen
---
v2:
- update copyrigh
the vector alignment changes introduced in GCC 5. Older gcc
versions forced vector alignment to 16 bytes due to requirements for
POWER 6 and earlier CPUs, but these restrictions don't apply to CPUs
supported by DPDK.
Bugzilla ID: 739
Signed-off-by: David Christensen
---
v2:
- update copyrigh
x86: ignore gcc 10 stringop-overflow warnings")
Signed-off-by: David Christensen
---
v2:
- update copyright year
- rebase for 21.11-rc0
---
lib/eal/ppc/include/rte_memcpy.h | 15 ---
1 file changed, 12 insertions(+), 3 deletions(-)
diff --git a/lib/eal/ppc/include/rte_memcpy.h b
failure. Modify the comparison to compare uint64_t values.
Fixes: fa3253c534b1 ("test/atomic: add 128-bit atomic compare exchange test")
Cc: phil.y...@arm.com
Signed-off-by: David Christensen
---
app/test/test_atomic.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
On 8/2/21 5:44 AM, Juraj Linkeš wrote:
+if os.name == 'posix':
+if os.path.isdir('/sys/devices/system/node'):
+numa_nodes = glob.glob('/sys/devices/system/node/node*')
+numa_nodes.sort()
+print(int(os.path.basename(numa_nodes[-1])[4:]) + 1)
+else:
+subpr
On 7/30/21 1:49 AM, jer...@marvell.com wrote:
From: Jerin Jacob
Dump the arm64 arch state register in oops
handling routine.
Signed-off-by: Jerin Jacob
---
lib/eal/unix/eal_oops.c | 19 +++
1 file changed, 19 insertions(+)
diff --git a/lib/eal/unix/eal_oops.c b/lib/eal/
On 7/30/21 1:49 AM, jer...@marvell.com wrote:
From: Jerin Jacob
Implement the base oops handling APIs.
Signed-off-by: Jerin Jacob
Building on POWER generates the following error:
ninja: Entering directory `build'
[1/244] Compiling C object 'lib/76b5a35@@rte_eal@sta/eal_unix_eal_oops.c.o
On 7/12/21 12:35 PM, David Christensen wrote:
On 7/5/21 2:28 AM, David Marchand wrote:
On Wed, Jun 23, 2021 at 6:13 PM David Christensen
wrote:
Suppress gcc warning "warning: writing 16 bytes into a region of
size 0" for users of the POWER rte_memcpy() function. Existing
On 7/16/21 6:53 AM, Juraj Linkeš wrote:
-Original Message-
From: David Christensen
Sent: Tuesday, July 6, 2021 8:11 PM
To: Bruce Richardson ; Juraj Linkeš
Cc: tho...@monjalon.net; david.march...@redhat.com;
honnappa.nagaraha...@arm.com; ruifeng.w...@arm.com;
ferruh.yi
On 7/5/21 2:28 AM, David Marchand wrote:
On Wed, Jun 23, 2021 at 6:13 PM David Christensen
wrote:
Suppress gcc warning "warning: writing 16 bytes into a region of
size 0" for users of the POWER rte_memcpy() function. Existing
rte_memcpy() code takes different code paths ba
On 6/15/21 1:12 AM, zhihongx.p...@intel.com wrote:
From: Zhihong Peng
AddressSanitizer (ASan) is a google memory error detect
standard tool. It could help to detect use-after-free and
{heap,stack,global}-buffer overflow bugs in C/C++ programs,
print detailed error information when error happ
On 7/6/21 2:08 AM, Bruce Richardson wrote:
On Tue, Jul 06, 2021 at 08:56:37AM +, Juraj Linkeš wrote:
-Original Message-
From: Bruce Richardson
Sent: Tuesday, June 29, 2021 1:29 PM
To: Juraj Linkeš
Cc: tho...@monjalon.net; david.march...@redhat.com;
honnappa.nagaraha...@arm.co
On 6/24/21 6:51 AM, David Marchand wrote:
On Tue, Jun 15, 2021 at 7:21 PM David Christensen
wrote:
Fix the IOMMU detection logic that looks for the "platform" field of
/proc/cpuinfo on POWER systems.
Fixes: 905215731833 ("bus/pci: support IOVA as VA on PowerNV sys
MMU : Hash
PowerNV KVM Guest $ cat /proc/cpuinfo
...
timebase : 51200
platform : pSeries
model: IBM pSeries (emulated by qemu)
machine : CHRP IBM pSeries (emulated by qemu)
MMU : Radix
Signed-off-by: David Christensen
Reviewed-by: Thinh Tran
---
drivers
Replacing memcpy() with rte_memcpy() improved 64 byte packet
performance by 33% on a POWER9 system and by 10% on an x86_64
system.
Signed-off-by: David Christensen
---
v2: Fixed last patch which was bungled due to a bad local git squash
drivers/net/memif/rte_eth_memif.c | 4 ++--
1 file
Replacing memcpy() with rte_memcpy() improved 64 byte packet
performance by 33% on a POWER9 system and by 10% on an x86_64
system.
Signed-off-by: David Christensen
---
drivers/net/memif/rte_eth_memif.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/net/memif
the vector alignment changes introduced in GCC 5. Older gcc
versions forced vector alignment to 16 bytes due to requirements for
POWER 6 and earlier CPUs, but these restrictions don't apply to CPUs
supported by DPDK.
Bugzilla ID: 739
Signed-off-by: David Christensen
---
config/ppc/meson
x86: ignore gcc 10 stringop-overflow warnings")
Signed-off-by: David Christensen
---
lib/eal/ppc/include/rte_memcpy.h | 13 +++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/lib/eal/ppc/include/rte_memcpy.h b/lib/eal/ppc/include/rte_memcpy.h
index c2a1f356d5..a84
Fix the IOMMU detection logic that looks for the "platform" field of
/proc/cpuinfo on POWER systems.
Fixes: 905215731833 ("bus/pci: support IOVA as VA on PowerNV systems")
Signed-off-by: David Christensen
---
drivers/bus/pci/linux/pci.c | 2 +-
1 file changed, 1 inse
get_tsc_freq_arch(void)
{
+#ifdef __GLIBC__
return __ppc_get_timebase_freq();
+#elif defined(__FreeBSD__)
+ uint64_t freq;
+ size_t length = sizeof(freq);
+ sysctlbyname("kern.timecounter.tc.timebase.frequency", &freq, &length,
NULL, 0);
+ return freq;
+#endif
}
Reviewed-by: David Christensen
On 5/5/21 3:39 PM, Piotr Kubaj wrote:
lib/librte_eal/ppc/include/rte_altivec.h | 3 +++
lib/librte_eal/ppc/include/rte_cycles.h | 12
lib/librte_eal/ppc/rte_cycles.c | 16
Directory path "librte_eal" has been changed to "eal" so a v3 patch is
like
er.tc.timebase.frequency", &freq, &length,
NULL, 0);
+ return freq;
+#endif
}
--
2.31.0
Reviewed-by: David Christensen
ld the lib names be specified with or without the prefix.
Therefore, we can just remove the library prefix and use the library's
unique name as the directory name, i.e. 'eal' rather than 'librte_eal'
Signed-off-by: Bruce Richardson
---
Reviewed-by: David Christensen
to eliminate the quick test timeout error.
Bugzilla ID: 684
Fixes: db1f2f8a9fe5 ("config: increase maximum lcores for ppc")
Cc: dwil...@us.ibm.com
Signed-off-by: David Christensen
---
config/ppc/meson.build | 17 ++---
1 file changed, 14 insertions(+), 3 deletions(-)
di
On 4/20/21 4:45 AM, luca.bocca...@gmail.com wrote:
From: Luca Boccassi
This test consistently times out on ppc64 builds. Disable it.
Cc: sta...@dpdk.org
Signed-off-by: Luca Boccassi
---
NAK. Will resolve with a different patch to reduce the max_lcore value
used for PPC builds. Both x
with -Dmax_lcores
and -Dmax_numa_nodes.
Signed-off-by: Juraj Linkeš
Reviewed-by: Honnappa Nagarahalli
---
Reviewed-by: David Christensen
orted
- machine = 'power8'
- machine_args = ['-mcpu=power8', '-mtune=power8']
- dpdk_conf.set('RTE_MACHINE','power8')
+machine = 'power8'
+machine_args = ['-mcpu=power8', '-mtune=power8']
+dpdk_conf.set('RTE_MACHINE','power8')
endif
# overrides specific to ppc64
Reviewed-by: David Christensen
On 4/20/21 4:45 AM, luca.bocca...@gmail.com wrote:
From: Luca Boccassi
This test consistently times out on ppc64 builds. Disable it.
Cc: sta...@dpdk.org
Signed-off-by: Luca Boccassi
Is there something new about how/when this test is run during continuous
integration? The test time is
First things first, let's Cc the PPC maintainer:
David Christensen
https://bugs.dpdk.org/show_bug.cgi?id=684
Tried GCC 10 on RHEL 8.3 and running the lcores_autotest individually
does not produce any errors. I can't see in the log file how the test
is called when it generate
vq->vq_free_cnt >= RTE_VIRTIO_VPMD_RX_REARM_THRESH) {
- virtio_rxq_rearm_vec(rxvq);
- if (unlikely(virtqueue_kick_prepare(vq)))
- virtqueue_notify(vq);
- }
-
nb_total = nb_used;
ref_rx_pkts = rx_pkts;
for (nb_pkts_received = 0;
Reviewed-by: David Christensen
On 3/31/21 2:06 AM, Juraj Linkeš wrote:
Add an option to automatically discover the host's numa and cpu counts
and use those values for a non cross-build.
Give users the option to override the per-arch default values or values
from cross files by specifying them on the command line with -Dmax_l
On 3/11/21 8:11 AM, Piotr Kubaj wrote:
There are couple of issues when building with Clang:
1. vector is a keyword and should not be used in code. I undefined it,
but it would probably be better to just change the variable name.
2. vector long is deprecated by Clang and should not be used. I s
On 2/19/21 1:11 AM, Juraj Linkeš wrote:
-Original Message-
From: Bruce Richardson
Sent: Wednesday, January 6, 2021 3:43 PM
To: David Christensen
Cc: Juraj Linkeš ; tho...@monjalon.net;
honnappa.nagaraha...@arm.com; dev@dpdk.org
Subject: Re: [RFC PATCH v2] build: add platform
Return a PPC specific value for get_tsc_freq_arch() rather than
depending on the EAL framework to estimate the frequency.
Signed-off-by: David Christensen
---
lib/librte_eal/ppc/rte_cycles.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/lib/librte_eal/ppc/rte_cycles.c b
On 1/18/21 7:27 PM, Xia, Chenbo wrote:
Also, when running vfio_user_autotest_server, I'm unable to exit the
application with CTRL-C directly. If a run a second test with
vfio_user_autotest_client then the server test runs to completion
without an error and I'm able to exit the test app normal
On 1/13/21 10:14 PM, Chenbo Xia wrote:
This patch introduces functional test for vfio_user client and
server. Note that the test can only be run with server and client
both started and server should be started first.
Receiving a build warning on RHEL 8.3 (gcc 8.3.1) for POWER with this patch
On 1/13/21 10:14 PM, Chenbo Xia wrote:
This patch introduces device and region related APIs, which are
rte_vfio_user_set_dev_info() and rte_vfio_user_set_reg_info().
The corresponding vfio-user command handling is also added with
the definition of all vfio-user command identity.
Receiving a
= RTE_IPV4(1, 0, 0, 0);
depth = 32;
- status = rte_lpm_add(lpm, ip, depth, next_hop_add);
+ status = rte_lpm_add(lpm, ip, depth, next_hop_base + ip);
TEST_LPM_ASSERT(status < 0);
rte_lpm_free(lpm);
Tested-by: David Christensen
i8.u32[3] = i8.u32[3] +
- (uint8_t)tbl[3] * RTE_LPM_TBL8_GROUP_NUM_ENTRIES;
+ (tbl[3] & 0x00FF) * RTE_LPM_TBL8_GROUP_NUM_ENTRIES;
ptbl = (const uint32_t *)&lpm->tbl8[i8.u32[3]];
tbl[3] = *ptbl;
}
Tested-by: David Christensen
-Original Message-
From: Thomas Monjalon
Sent: Friday, January 8, 2021 11:09 AM
To: Ophir Munk
Cc: dev@dpdk.org; Ori Kam
Subject: Re: [dpdk-dev] [PATCH v2 5/6] app/regex: support performance
measurements per QP
20/12/2020 11:41, Ophir Munk:
Up to this commit measuring the parsing elap
1 - 100 of 288 matches
Mail list logo
