[dpdk-dev] app/testpmd:fix invalid port id parameters
Checks
Commit Message
in parse_ringnuma_config/parse_portnuma_config functions,port_id
should less than RTE_MAX_ETHPORTS,but port_id_is_invalid check
assumes that port_id may be 255.
Signed-off-by: Li Han <han.li1@zte.com.cn>
---
app/test-pmd/parameters.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
Comments
Hi Han,
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Li Han
> Sent: Tuesday 22 August 2017 06:04
> To: Wu, Jingjing <jingjing.wu@intel.com>
> Cc: dev@dpdk.org; Li Han <han.li1@zte.com.cn>
> Subject: [dpdk-dev] [PATCH] app/testpmd:fix invalid port id parameters
>
> in parse_ringnuma_config/parse_portnuma_config functions,port_id should
> less than RTE_MAX_ETHPORTS,but port_id_is_invalid check assumes that
> port_id may be 255.
>
> Signed-off-by: Li Han <han.li1@zte.com.cn>
> ---
> app/test-pmd/parameters.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/app/test-pmd/parameters.c b/app/test-pmd/parameters.c index
> 2f7f70f..0c97ba4 100644
> --- a/app/test-pmd/parameters.c
> +++ b/app/test-pmd/parameters.c
> @@ -424,7 +424,8 @@
> return -1;
> }
> port_id = (uint8_t)int_fld[FLD_PORT];
> - if (port_id_is_invalid(port_id, ENABLED_WARN)) {
> + if (port_id_is_invalid(port_id, ENABLED_WARN) ||
> + port_id == (portid_t)RTE_PORT_ALL) {
In this case the message "printf("Invalid port %d\n", port_id)" will be omitted and there will be no way
To figure out which port id caused the problem
> printf("Valid port range is [0");
> RTE_ETH_FOREACH_DEV(pid)
> printf(", %d", pid);
> @@ -483,7 +484,8 @@
> return -1;
> }
> port_id = (uint8_t)int_fld[FLD_PORT];
> - if (port_id_is_invalid(port_id, ENABLED_WARN)) {
> + if (port_id_is_invalid(port_id, ENABLED_WARN) ||
> + port_id == (portid_t)RTE_PORT_ALL) {
The same here
> printf("Valid port range is [0");
> RTE_ETH_FOREACH_DEV(pid)
> printf(", %d", pid);
> --
> 1.8.3.1
>
Thanks,
Kirill
> -----Original Message-----
> From: Rybalchenko, Kirill
> Sent: Monday, September 4, 2017 10:12 PM
> To: Li Han <han.li1@zte.com.cn>; Wu, Jingjing <jingjing.wu@intel.com>
> Cc: dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH] app/testpmd:fix invalid port id parameters
>
> Hi Han,
>
> > -----Original Message-----
> > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Li Han
> > Sent: Tuesday 22 August 2017 06:04
> > To: Wu, Jingjing <jingjing.wu@intel.com>
> > Cc: dev@dpdk.org; Li Han <han.li1@zte.com.cn>
> > Subject: [dpdk-dev] [PATCH] app/testpmd:fix invalid port id parameters
> >
> > in parse_ringnuma_config/parse_portnuma_config functions,port_id should
> > less than RTE_MAX_ETHPORTS,but port_id_is_invalid check assumes that
> > port_id may be 255.
> >
> > Signed-off-by: Li Han <han.li1@zte.com.cn>
> > ---
> > app/test-pmd/parameters.c | 6 ++++--
> > 1 file changed, 4 insertions(+), 2 deletions(-)
> >
> > diff --git a/app/test-pmd/parameters.c b/app/test-pmd/parameters.c index
> > 2f7f70f..0c97ba4 100644
> > --- a/app/test-pmd/parameters.c
> > +++ b/app/test-pmd/parameters.c
> > @@ -424,7 +424,8 @@
> > return -1;
> > }
> > port_id = (uint8_t)int_fld[FLD_PORT];
> > - if (port_id_is_invalid(port_id, ENABLED_WARN)) {
> > + if (port_id_is_invalid(port_id, ENABLED_WARN) ||
> > + port_id == (portid_t)RTE_PORT_ALL) {
> In this case the message "printf("Invalid port %d\n", port_id)" will be omitted and there
> will be no way
> To figure out which port id caused the problem
Why? If the port_id is invalid and not RTE_PORT_ALL, the "printf("Invalid port %d\n", port_id)" should be print.
On 8/22/2017 6:03 AM, Li Han wrote:
> in parse_ringnuma_config/parse_portnuma_config functions,port_id
> should less than RTE_MAX_ETHPORTS,but port_id_is_invalid check
> assumes that port_id may be 255.
testpmd using RTE_PORT_ALL [1], which is in valid port_id range, as
special meaning [2] making things tricky.
Only above 255 is no more valid since port_id is not 16bits, it should
be 65535.
Except from above detail,
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
And if you are familiar with this code, above a few lines there is a
"int_fld[i] > 255" check, is the intention to check port_id limit there?
If so this is no more valid check, and would you mind sending a patch to
fix if you have time?
Thanks,
ferruh
[1]
#define RTE_PORT_ALL (~(portid_t)0x0)
[2]
Meaning all enabled ethdev ports, like:
start_port(RTE_PORT_ALL);
>
> Signed-off-by: Li Han <han.li1@zte.com.cn>
<...>
On 10/9/2017 5:16 AM, Ferruh Yigit wrote:
> On 8/22/2017 6:03 AM, Li Han wrote:
>> in parse_ringnuma_config/parse_portnuma_config functions,port_id
>> should less than RTE_MAX_ETHPORTS,but port_id_is_invalid check
>> assumes that port_id may be 255.>>
>> Signed-off-by: Li Han <han.li1@zte.com.cn>
> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Applied to dpdk-next-net/master, thanks.
Welcome Li Han!
(Fixed mentioned 255 note in commit log while applying, also fixed local
port_id storage size while applying)
@@ -424,7 +424,8 @@
return -1;
}
port_id = (uint8_t)int_fld[FLD_PORT];
- if (port_id_is_invalid(port_id, ENABLED_WARN)) {
+ if (port_id_is_invalid(port_id, ENABLED_WARN) ||
+ port_id == (portid_t)RTE_PORT_ALL) {
printf("Valid port range is [0");
RTE_ETH_FOREACH_DEV(pid)
printf(", %d", pid);
@@ -483,7 +484,8 @@
return -1;
}
port_id = (uint8_t)int_fld[FLD_PORT];
- if (port_id_is_invalid(port_id, ENABLED_WARN)) {
+ if (port_id_is_invalid(port_id, ENABLED_WARN) ||
+ port_id == (portid_t)RTE_PORT_ALL) {
printf("Valid port range is [0");
RTE_ETH_FOREACH_DEV(pid)
printf(", %d", pid);