The lcore variables API is new in DPDK 24.11,
that's why the function rte_lcore_var_alloc() was marked experimental.
To be clearer, the whole header file (including all macros)
is marked experimental.
A change decreased the default buffer size from 1 MB to 128 kB,
missing to update the documentati
+1
What about using thread id instead?
From d1687ffbf865ba0b2d64c35acd602ca43329691e Mon Sep 17 00:00:00 2001
From: Stephen Hemminger
Date: Thu, 14 Nov 2024 08:48:54 -0800
Subject: [PATCH] eal: fix fbarray name with multiple secondary processes
When multiple secondary processes run in different con
Fallthrough warnings catch real bugs and should not be disabled.
There are warnings in this driver in current build.
The commit that added the disable is old, and the problematic code
appears to have been already removed.
Fixes: 2ab9a9483196 ("crypto/dpaa2_sec: fix build with GCC 7")
Cc: tho...@m
Fix bug detected from bad fallthroug in E1000 code.
And re-enable compiler flag in both e1000 and dpaax
v2 - minor comment fix
Stephen Hemminger (2):
common/dpaax: do not disable fallthrough warnings
net/e1000: fix incorrect fallthrough in switch
drivers/common/dpaax/caamflib/rta/operation_
On Thu, 14 Nov 2024 14:38:05 +0530
wrote:
>
> +uint8_t
> +graph_is_node_active_in_graph(struct node *node)
For true/false functions bool type is preferred.
It prevents cases where function accidentally returns 2.
There is an incorrect fallthrough identified by PVS studio.
Even though this is in base code it should be fixed, and
the warning should be re-enabled to prevent future bugs.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: f2553cb9eba6 ("net/e1000/base: add new I219 devices")
Cc: qimin
Usual problem with the commit subject.
On 23/09/2024 19:42, jspew...@iol.unh.edu wrote:
+def set_num_virtual_functions(self, num: int, pf_port: Port) -> bool:
I'd call this set_num_of_virtual_functions, set_virtual_functions_num,
or set_number_of_vfs. I think I prefer the last one.
+
On 23/09/2024 19:42, jspew...@iol.unh.edu wrote:
+
+def get_vfs_on_port(self, pf_port: Port) -> list[VirtualFunction]:
+"""Get all virtual functions (VFs) that DTS is aware of on `pf_port`.
+
+Args:
+pf_port: The port to search for the VFs on.
+
+Returns:
+
On Thu, 14 Nov 2024 14:38:06 +0530
wrote:
> From: Kiran Kumar K
>
> The graph autotest second run test is failing due to the
> node name is already present in the node list. Adding changes
> to free nodes at the time of test cleanup.
>
> Fixes: 6b89650418fe ("test/graph: add functional tests")
https://bugs.dpdk.org/show_bug.cgi?id=1581
Bug ID: 1581
Summary: net/txgbe: duplicate legs of if statement
Product: DPDK
Version: 24.11
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: major
P
The same condition is checked twice in an if statement.
Harmless, but redundant.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Signed-off-by: Stephen Hemminger
---
drivers/common/cnxk/cnxk_security.c | 16 ++--
1 file changed, 10 insertions(+), 6 deletions(-)
diff --git a/dr
The dqrr_held mask is 64 bit but updates were getting truncated
because 1 is of type int (32 bit) and the result shift of int is of
type int (32 bit); therefore any value >= 32 would get truncated.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: a77db24643b7 ("crypto/dpaa2_sec: suppor
The log message would always refer to str variable which
is NULL here. Looks like author intended to print original
parameter.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Signed-off-by: Stephen Hemminger
---
drivers/net/cpfl/cpfl_ethdev.c | 2 +-
1 file changed, 1 insertion(+), 1 delet
https://bugs.dpdk.org/show_bug.cgi?id=1579
Bug ID: 1579
Summary: [dpdk-24.11-rc2] meson_tests/fasts malloc_autotest
test failed on i40e
Product: DPDK
Version: 24.11
Hardware: All
OS: All
Status:
The dqrr_held mask is 64 bit but updates were getting truncated
because 1 is of type int (32 bit) and the result shift of int is of
type int (32 bit); therefore any value >= 32 would get truncated.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: 2d3788631862 ("net/dpaa2: support atomi
The function would leak file if fscanf failed.
There is a working version in other file, clone that.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: 0e6557b448fa ("raw/cnxk_gpio: add self test")
Cc: tduszyn...@marvell.com
Signed-off-by: Stephen Hemminger
---
drivers/raw/cnxk_gpio/c
The type mask is 64 bit value, doing a shift of literal 1 (32 bit)
will result in int type (32 bit) and cause truncation.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: f4ca3fd54c4d ("net/hinic: create and destroy flow director filter")
Cc: cloud.wangxiao...@huawei.com
Cc: sta...@dpdk
More bitmask truncation from mask computation.
Fixes: 0ee17f79ebd0 ("event/dpaa: add enqueue/dequeue")
Cc: sunil.k...@nxp.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hemminger
---
drivers/event/dpaa/dpaa_eventdev.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git
> The same condition is checked twice in an if statement.
> Harmless, but redundant.
>
> Link:
> https://urldefense.proofpoint.com/v2/url?u=https-3A__pvs-2Dstudio.com_en_blog_posts_cpp_1183_&d=DwIDAg&c=nKjWec2b6R0mOyPaz7xtfQ&r=jPfB8rwwviRSxyLWs2n6B-
> WYLn1v9SyTMrT5EQqh2TU&m=tCMuFtN3iZvm6NW4QbOzKC
On Thu, 14 Nov 2024, Stephen Hemminger wrote:
>
> In general DPDK uses tsc instead of monotonic time, since it is faster and
> platform
> independent (ie Windows).
>
> Why not use use a global counter instead?
>
I have tried tsc, but it only works on x86 and does not function on other
archi
On Thu, 14 Nov 2024, Stephen Hemminger wrote:
> Date: Fri, 15 Nov 2024 01:06:27
> From: Stephen Hemminger
> To: zcjie0802
> Cc: dev , anatoly.burakov
> Subject: Re: [PATCH] eal/linux: redefine the name for rte_fbarray_init()
>
> What about using thread id instead?
>
> From d1687ffbf865ba0b2d6
When memset() is used before a release function such as free,
the compiler if allowed to optimize the memset away under
the as-if rules. This is normally ok, but in certain cases such
as passwords or security keys it is problematic.
Introduce a DPDK wrapper which is equivalent to the C++ memset_s
Calling memset before free() has no effect and will be flagged
by security parsing tools as a potential bug. None of these data
structures have sensitive information.
Signed-off-by: Stephen Hemminger
---
drivers/net/ntnic/nthw/core/nthw_hif.c| 5 +
drivers/net/ntnic/nthw/core
Need to check the result of malloc() before calling memset.
This is only place in this driver that forgot, other code
does check.
Signed-off-by: Stephen Hemminger
---
drivers/net/ntnic/nthw/nthw_rac.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ntnic/nthw/n
Applied to next-dts with commit message updated, thanks.
Applied to next-dts with commit message updated, thanks.
Applied to next-dts with commit message updated, thanks.
The members "dispatch" and "xstat_off" of the structure "rte_node"
can be min cache aligned to make room for future expansion and to
make sure have better performance. Add corresponding comments.
Signed-off-by: Huichao Cai
---
doc/guides/rel_notes/release_24_11.rst | 2 ++
lib/graph/rte_graph_w
On Thu, 14 Nov 2024, Stephen Hemminger wrote:
>
> In general DPDK uses tsc instead of monotonic time, since it is faster and
> platform
> independent (ie Windows).
>
> Why not use use a global counter instead?
>
I have tried tsc, but it only works on x86 and does not function on other archs.
https://bugs.dpdk.org/show_bug.cgi?id=1580
Bug ID: 1580
Summary: Nthw: array overrun
Product: DPDK
Version: 24.11
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: Normal
On Fri, 15 Nov 2024 10:26:00 +0800
Zhou congjie wrote:
> On Thu, 14 Nov 2024, Stephen Hemminger wrote:
>
> > Date: Fri, 15 Nov 2024 01:06:27
> > From: Stephen Hemminger
> > To: zcjie0802
> > Cc: dev , anatoly.burakov
> > Subject: Re: [PATCH] eal/linux: redefine the name for rte_fbarray_init()
Doing memset before free maybe removed by compiler, and
is flagged by security scanning tools as potential problem.
In this case the memset is unnecessary.
Signed-off-by: Stephen Hemminger
---
drivers/bus/uacce/uacce.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/bus/uacce/uacce.c
Regular memset maybe removed by compiler if done before a free
function. Use new rte_free_sensitive instead.
Signed-off-by: Stephen Hemminger
Acked-by: Bruce Richardson
---
drivers/crypto/qat/qat_asym.c | 5 +
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/drivers/crypto/qat/q
Series-acked-by: Chengwen Feng
On 2024/11/14 19:39, Dengdui Huang wrote:
> Whether strerror() needs to be replaced needs further discussion,
> but rte_error should not be used by strerror() because rte_errno
> may be an RTE-specific error code.
>
> This patchset use rte_strerror() instead of st
There is no sensitive information in dev, so
Acked-by: Chengwen Feng
On 2024/11/15 2:43, Stephen Hemminger wrote:
> Doing memset before free maybe removed by compiler, and
> is flagged by security scanning tools as potential problem.
> In this case the memset is unnecessary.
>
> Signed-off-by: S
More bug fixes from PVS studio bug reports.
And one other fix to ptpclient.
Stephen Hemminger (16):
common/cnxk: remove duplicate condition
net/cpfl: avoid calling log (printf) with null
raw/cnxk_gpio: fix file descriptor leak
net/ntnic: remove dead code
net/i40e: remove duplicate code
The loop to update speed would not work because num_port_speeds
was always zero so it did nothing. And the array of pls_mbps
was only used inside the loop but never set.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Signed-off-by: Stephen Hemminger
---
drivers/net/ntnic/ntnic_ethdev.c |
There are two branches in the cascading if/else that have same
condition and code; remove one. Update the code to follow DPDK
style where all statements in if should have brackets if any
leg requires them.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Signed-off-by: Stephen Hemminger
---
The dqrr_held mask is 64 bit but updates were getting truncated
because 1 is of type int (32 bit) and the result shift of int is of
type int (32 bit); therefore any value >= 32 would get truncated.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: 5e7455931442 ("net/dpaa: support Rx que
Calling memcmp on same structure will always be true.
Replace with same conditional used elsewhere.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: ab129e9065a5 ("examples/ptpclient: add minimal PTP client")
Cc: danielx.t.mrzyg...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen
The dqrr_held mask is 64 bit but updates were getting truncated
because 1 is of type int (32 bit) and the result shift of int is of
type int (32 bit); therefore any value >= 32 would get truncated.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: fe3688ba7950 ("crypto/dpaa_sec: support
The array of debug status strings did not match possible enum
values. Add the missing element and a static assert to make sure
the table has all possible values.
For more complete description see.
Link: https://pvs-studio.com/en/blog/posts/cpp/1176/
Fixes: ec55c118792b ("net/qede: add infrastruct
Both sides of the if in uninit are using same code.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Signed-off-by: Stephen Hemminger
---
drivers/net/octeon_ep/otx_ep_ethdev.c | 9 ++---
1 file changed, 2 insertions(+), 7 deletions(-)
diff --git a/drivers/net/octeon_ep/otx_ep_ethdev.c
Don't use rte_memcpy() when not necessary. Structure assignment
is as fast and type safe.
Signed-off-by: Stephen Hemminger
---
examples/ptpclient/ptpclient.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/examples/ptpclient/ptpclient.c b/examples/ptpclient/ptpclient.
The code for parsing layers in devargs could reference past
the end of layers[] array on stack.
Link: https://pvs-studio.com/en/blog/posts/cpp/1183/
Fixes: 9a1a9e4a2ddd ("devargs: support path value with global device syntax")
Cc: xuemi...@nvidia.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hem
On 2024/11/14 8:12, Stephen Hemminger wrote:
> There was useless loop when looking at the DMA address.
> It looks like it was meant to skip whitespace before
> calling strtok.
>
> Good time to replace strtok with strtok_r as well.
Incomplete modification for strtok, I suggest the strtok adopt Hai
add the tsc into the name.
Suggested-by: Stephen Hemminger
Signed-off-by: Congjie Zhou
---
When multiple secondary processes run in different containers, names
identified by PIDs are not unique due to the pid namespace.
So Add tsc to redefine a unique name.
v1: use monotonic time to redefine
Script that converts memset before free into rte_memset_sensitive
and memset before rte_free into rte_free_sensitive
Signed-off-by: Stephen Hemminger
---
devtools/cocci/memset_free.cocci | 20
1 file changed, 20 insertions(+)
create mode 100644 devtools/cocci/memset_free.co
14/11/2024 15:12, Serhii Iliushyk:
> >On 11.11.2024, 15:35, "Ferruh Yigit" wrote:
> >
> >
> >On 10/31/2024 3:00 PM, Danylo Vodopianov wrote:
> >> These patches provide next fixes:
> >> 1. The testpmd command “flow update…“ provides a nullptr as the
> >> context variable.
> >> 2. Avoid removal of ad
The memset was always doing 0 bytes since size computed later.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 3a80d7fb2ecd ("crypto/qat: support SHA3 plain hash")
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hemminger
Acked-by: Bruce Richardson
---
drivers/crypto/qat/qat_sym_sessio
There is no need to cast memset to void.
Signed-off-by: Stephen Hemminger
---
drivers/net/ntnic/nim/i2c_nim.c | 2 +-
drivers/net/ntnic/nthw/flow_filter/flow_nthw_cat.c| 4 ++--
drivers/net/ntnic/nthw/flow_filter/flow_nthw_csu.c| 4 ++--
drivers/net/ntnic/nthw/flow_
Although internally rte_free does poison the buffer in most
cases, it is useful to have function that explicitly does
this to avoid any security issues.
Signed-off-by: Stephen Hemminger
---
lib/eal/common/rte_malloc.c | 30 --
lib/eal/include/rte_malloc.h | 18 ++
This series handles memset related bugs indentified by PVS Studio.
The root cause is that Gcc and other compilers are free to
optimize away memset called before free.
Most of the places memset was being used like this were bogus;
probably some developer debug habit, and can be safely removed.
v4
Just doing memset() on keys is not enough, compiler can optimize
it away. Need something with a barrier.
Signed-off-by: Stephen Hemminger
Acked-by: Bruce Richardson
---
drivers/crypto/qat/qat_sym_session.c | 17 +
1 file changed, 9 insertions(+), 8 deletions(-)
diff --git a/dri
The test loop was much shorter than desired because when
MAX_NUM is defined with out paren's the divide operator /
takes precedence over shift.
But when MAX_NUM is fixed, some tests take too long
and have to modified to avoid running over full N^2
space of 1<<20.
Note: this is a very old bug, goe
Do not use same variable for outer and inner loop in bonding test.
Since the loop is just freeing the resulting burst use bulk free.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 92073ef961ee ("bond: unit tests")
Cc: declan.dohe...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Steph
The order of comparison is wrong, and potentially allows
referencing past the array.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 3e3edab530a1 ("ethdev: add flow quota")
Cc: getel...@nvidia.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hemminger
---
app/test-pmd/cmdline_flow.c
On Thu, Nov 14, 2024 at 11:26 AM Stephen Hemminger
wrote:
>
> The loop over policy actions will always exit when it sees
> the flow end action, so the next check is redundant.
>
> Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
>
> Fixes: f29fa2c59b85 ("app/testpmd: support policy actions per
Since tmp is not used later in the function, this memset
is unnecessary. Even though this is harmless,
it causes tools that look for security issues
around memset to flag this a bug.
Signed-off-by: Stephen Hemminger
Acked-by: Bruce Richardson
---
app/test/test_cmdline_cirbuf.c | 2 --
1 file ch
Hi there,
once again I am in agreement with Juraj, this approach may risk to
complicate things further. As he mentioned, we should allow TestSuite to
be marked for VFs and they should be able to automate the setup behind
the scenes. If we wanted to distinguish VF and PF test cases, I think
th
Calling memset before rte_free not necessary, and could be
removed by the compiler. In this case, the data is not security
sensitive so the memset can be removed. Some security scanning
tools will flag this.
Signed-off-by: Stephen Hemminger
Acked-by: Bruce Richardson
---
drivers/compress/octeon
Recent blog post from PVS studio referenced lots bugs found by
their analyzer against DPDK. This set addresses the ones in
the test suite.
v2 - fix dma address parsing
- add stable where needed
Stephen Hemminger (10):
app/test: do not duplicated loop variable
app/test: fix typo in address
The first argument of 'memcmp' function was equal to the second argument.
Therefore ASSERT would always be true.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 92073ef961ee ("bond: unit tests")
Cc: declan.dohe...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hemminger
---
a
The event_dev_config initialization had duplicate assignments
to the same element. Change to use structure initialization
so that compiler will catch this type of bug.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: f8f9d233ea0e ("test/eventdev: add unit tests")
Cc: jerin.ja...@cavium
The code was duplicating the same condition three times?
Reading the commit message, the intention was:
Add unit tests to verify the zero len TLS records. Zero len packets are
allowed when content type is app data while zero packet length with
other content type (such as handshake) wou
The parenthesis were in the wrong place so that comparison
took precedence over assignment in handling IPv6 extension
headers. Break up the loop condition to avoid the problem.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 15ccc647526e ("test/security: test inline reassembly with m
The expression for checking which lcore is enabled for 0-7
was wrong (missing case for 6).
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: b0209034f2bb ("test/eal: check number of cores before running subtests")
Cc: msant...@redhat.com
Cc: sta...@dpdk.org
Signed-off-by: Stephen Hemmi
There was useless loop when looking at the DMA address.
It looks like it was meant to skip whitespace before
calling strtok.
Good time to replace strtok with strtok_r as well.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test"
The loop over policy actions will always exit when it sees
the flow end action, so the next check is redundant.
Link: https://pvs-studio.com/en/blog/posts/cpp/1179/
Fixes: f29fa2c59b85 ("app/testpmd: support policy actions per color")
Cc: haif...@nvidia.com
Signed-off-by: Stephen Hemminger
---
https://bugs.dpdk.org/show_bug.cgi?id=1578
Bug ID: 1578
Summary: Observed Inconsistencies In Device Allow-list Pools.
Product: DPDK
Version: unspecified
Hardware: All
OS: All
Status: UNCONFIRMED
Severity:
Applied to next-dts, thanks.
add the current time into the name
Signed-off-by: Congjie Zhou
---
When multiple secondary processes run in different containers, names
identified by PIDs are not unique due to the pid namespace.
Add current time to redefine a unique name.
lib/eal/linux/eal_memalloc.c | 16 ++--
1
add the current time into the name
Signed-off-by: Congjie Zhou
---
When multiple secondary processes run in different containers, names
identified by PIDs are not unique due to the pid namespace.
Add current time to redefine a unique name.
lib/eal/linux/eal_memalloc.c | 17 +++--
1
Add queue initialization and release interface.
Signed-off-by: Yong Zhang
---
drivers/raw/gdtc/gdtc_rawdev.c | 244 +
drivers/raw/gdtc/gdtc_rawdev.h | 19 +++
2 files changed, 263 insertions(+)
diff --git a/drivers/raw/gdtc/gdtc_rawdev.c b/drivers/raw/gdtc/gdtc_
Introduce rawdev driver support for GDTC which
can help to connect two separate hosts with each other.
Signed-off-by: Yong Zhang
---
.mailmap | 1 +
MAINTAINERS| 5 +
doc/guides/rawdevs/gdtc.rst| 35 ++
doc/guides/rawdevs/index.rst | 1 +
Release status meeting minutes 2024-11-14
=
Agenda:
* Release Dates
* Subtrees
* Roadmaps
* LTS
* Defects
* Opens
Participants:
* AMD
* ARM
* Debian/Microsoft
* Intel
* Marvell
* Nvidia
* Red Hat
Release Dates
-
The following are the current/up
> -Original Message-
> From: Huichao Cai
> Sent: Thursday, November 14, 2024 2:15 PM
> To: Jerin Jacob ; Kiran Kumar Kokkilagadda
> ; Nithin Kumar Dabilpuram
> ; yanzhirun_...@163.com
> Cc: dev@dpdk.org
> Subject: [EXTERNAL] [PATCH v4 2/2] graph: add alignment to the member of
> rte_nod
Hi,
> wrote:
> >
> > Add co-maintainer for DMA perf app.
> >
> > Signed-off-by: Gowrishankar Muthukrishnan
>
> Good to have more maintainers.
>
> Acked-by: Jerin Jacob
>
>
Could this patch be applied ?.
Thanks,
Gowrishankar
>
>
> > ---
> > MAINTAINERS | 1 +
> > 1 file changed, 1 insert
Although it should never occur that the teid (scheduler node id) value
for a txq is incorrect, put in a check for null following the scheduler
node lookup in ice_tm_setup_txq_node. This provides some additional
safety and should eliminate a coverity issue too.
Coverity issue: 448957
Fixes: 715d449
Hi, Jerin. Like this?
diff --git a/lib/graph/rte_graph_worker_common.h
b/lib/graph/rte_graph_worker_common.h
index a518af2b2a..f9ff7dd8c9 100644
--- a/lib/graph/rte_graph_worker_common.h
+++ b/lib/graph/rte_graph_worker_common.h
@@ -104,15 +104,19 @@ struct __rte_cache_aligned rte_node {
Hi Dean,
This looks fine for most of it, but some of Juraj's prior comments still
apply. See comments in line.
On 10/10/2024 22:06, Dean Marx wrote:
+@dataclass
I'd make this kw_only=True:
@dataclass(kw_only=True)
this enforces the construction to include field names for the sake of
re
> -Original Message-
> From: Huichao Cai
> Sent: Thursday, November 14, 2024 5:37 PM
> To: Jerin Jacob
> Cc: Kiran Kumar Kokkilagadda ; Nithin Kumar
> Dabilpuram ; yanzhirun_...@163.com;
> david.march...@redhat.com; dev@dpdk.org
> Subject: Re:RE: [EXTERNAL] [PATCH v4 2/2] graph: add al
Hi everyone,
Looking at implementation of cryptodev callbacks
(it uses DPDK RCU), it seems like there is a bug here:
at init time we don't call rte_rcu_qsbr_thread_register().
As I understand without it rte_rcu_qsbr_check() wouldn't
work properly for that thread.
Probably need to add:
static int
From: Kiran Kumar K
Add support for rte_node_free API to free the node and its
memory, if node is not part of any of the created graphs.
Signed-off-by: Kiran Kumar K
---
lib/graph/graph.c | 16
lib/graph/graph_private.h | 13 +
lib/graph/node.c | 2
Hi folks,
Morten Brørup, Nov 14, 2024 at 08:43:
Medvedkin, Vladimir:
I think control plane API should work with prefix addresses in CPU
byte order. At least our RTE_IPV4 macro works this way. Also, prefix
is an address + prefix length (not the mask), so it is more natural
if address is in cpu
Add rawdev dequeue operation for gdtc devices.
Signed-off-by: Yong Zhang
---
drivers/raw/gdtc/gdtc_rawdev.c | 168 -
1 file changed, 147 insertions(+), 21 deletions(-)
diff --git a/drivers/raw/gdtc/gdtc_rawdev.c b/drivers/raw/gdtc/gdtc_rawdev.c
index f58b034e58..
Found couple of reasons causing mailbox timeout:
* msgs_acked value by interrupt thread was not getting synced to thread
polling it.
* mbox_data value sent by AF was seen 0 causing neither down nor up msg
processing.
Fixes: 9bd368ca311a ("common/cnxk: enable PF VF mailbox")
Fixes: 61deac72abbf ("c
Implementing mailbox debug trace points to trace function calls.
Signed-off-by: Harman Kalra
---
V2:
* Fixed line length exceeding 100 columns
drivers/common/cnxk/roc_dev.c | 87 --
drivers/common/cnxk/roc_mbox.c | 34
drivers/common/cnxk/roc_
Add support for rawdev operations such as dev_start and dev_stop.
Signed-off-by: Yong Zhang
---
drivers/raw/gdtc/gdtc_rawdev.c | 136 -
drivers/raw/gdtc/gdtc_rawdev.h | 10 +++
2 files changed, 145 insertions(+), 1 deletion(-)
diff --git a/drivers/raw/gdtc/gdtc_
Hi Thomas,
>> I submitted a patch on October 29 and haven't received response.>> Wondering
>> if any modifications are needed? Could you provide an update on the
>> review?>> Thanks. Looking forward to hearing from you.>>I cannot apply your
>> patches because the format looks wrong.>It looks th
Representor port link status should be toggled based on representee
state i.e. if representee comes up representor link status should be
up while when representee goes down representor port should indicate
the same.
Signed-off-by: Harman Kalra
---
drivers/net/cnxk/cnxk_rep_ops.c | 26 +++
From: Kiran Kumar K
The node id is determined based on a global variable that is
incremented every time a node is created. Adding changes to
remove the global counter. Make sure that the node list is always
ordered by increasing node ids. When creating a new node, pick a free
id which is not allo
From: Kiran Kumar K
Add support for rte_node_free API to free the node and its
memory, if node is not part of any of the created graphs.
Signed-off-by: Kiran Kumar K
---
lib/graph/graph.c | 16
lib/graph/graph_private.h | 13 +
lib/graph/node.c | 2
From: Kiran Kumar K
The graph autotest second run test is failing due to the
node name is already present in the node list. Adding changes
to free nodes at the time of test cleanup.
Fixes: 6b89650418fe ("test/graph: add functional tests")
Signed-off-by: Kiran Kumar K
---
app/test/test_graph.c
On Wed, Nov 13, 2024 at 04:36:13PM -0800, Stephen Hemminger wrote:
> There is an incorrect fallthrough identified by PVS studio.
> Even though this is in base code it should be fixed, and
> the warning should be re-enabled to prevent future bugs.
>
> Link: https://pvs-studio.com/en/blog/posts/cpp/
In the function __rte_graph_mcore_dispatch_sched_node_enqueue,
use a slower loop to search for the graph, modify the search logic
to record the result of the first search, and use this record for
subsequent searches to improve search speed.
Due to the addition of a "graph" field in the "rte_node"
The members dispatch and xstat_off of the structure rte_node
can be min cache aligned to make room for future expansion and to
make sure have better performance. Add corresponding comments.
Due to the modification of the alignment of some members of the
rte_node structure, update file release_24_1
On Wed, Nov 13, 2024 at 10:55:59AM -0800, Stephen Hemminger wrote:
> Bruce mentioned recent security scan of DPDK
> https://pvs-studio.com/en/blog/posts/cpp/1183/
>
> There are some low hanging fruit things that can be easily addressed.
>
> Stephen Hemminger (5):
> crypto/qat: fix memset
Add rawdev enqueue operation for gdtc devices.
Signed-off-by: Yong Zhang
---
drivers/raw/gdtc/gdtc_rawdev.c | 220 +
drivers/raw/gdtc/gdtc_rawdev.h | 19 +++
2 files changed, 239 insertions(+)
diff --git a/drivers/raw/gdtc/gdtc_rawdev.c b/drivers/raw/gdtc/gdtc_r
1 - 100 of 124 matches
Mail list logo
