Re: [dpdk-dev] [PATCH v3] eal: fix core number validation

[David Marchand <david.marchand@redhat.com>]

Hello Hari,

On Mon, Jan 7, 2019 at 11:26 AM Hari Kumar Vemula <
hari.kumarx.vemula@intel.com> wrote:

> When incorrect core value or range provided,
> as part of -l command line option, a crash occurs.
>
> Added valid range checks to fix the crash.
>
> Added ut check for negative core values.
> Added unit test case for invalid core number range.
>
> Fixes: d888cb8b9613 ("eal: add core list input format")
> Cc: stable@dpdk.org
>
> --
> v3: Added unit test cases for invalid core number range
> v2: Replace strtoul with strtol
>     Modified log message
> --
>
> Signed-off-by: Hari Kumar Vemula <hari.kumarx.vemula@intel.com>
> ---
>  lib/librte_eal/common/eal_common_options.c |  9 +++++++--
>  test/test/test_eal_flags.c                 | 14 +++++++++++++-
>  2 files changed, 20 insertions(+), 3 deletions(-)
>
> diff --git a/lib/librte_eal/common/eal_common_options.c
> b/lib/librte_eal/common/eal_common_options.c
> index 6e3a83b98..9431c024e 100644
> --- a/lib/librte_eal/common/eal_common_options.c
> +++ b/lib/librte_eal/common/eal_common_options.c
> @@ -592,7 +592,9 @@ eal_parse_corelist(const char *corelist)
>                 if (*corelist == '\0')
>                         return -1;
>                 errno = 0;
> -               idx = strtoul(corelist, &end, 10);
> +               idx = strtol(corelist, &end, 10);
> +                       if (idx < 0 || idx >= (int)cfg->lcore_count)
> +                               return -1;
>                 if (errno || end == NULL)
>                         return -1;
>                 while (isblank(*end))
>

Please fix indentation.


> @@ -1103,6 +1105,7 @@ eal_parse_common_option(int opt, const char *optarg,
>  {
>         static int b_used;
>         static int w_used;
> +       struct rte_config *cfg = rte_eal_get_configuration();
>
>         switch (opt) {
>         /* blacklist */
> @@ -1145,7 +1148,9 @@ eal_parse_common_option(int opt, const char *optarg,
>         /* corelist */
>         case 'l':
>                 if (eal_parse_corelist(optarg) < 0) {
> -                       RTE_LOG(ERR, EAL, "invalid core list\n");
> +                       RTE_LOG(ERR, EAL,
> +                               "Invalid core number, core range should be
> (0, %u)\n",
> +                                       cfg->lcore_count-1);
>                         return -1;
>                 }
>
>
eal_parse_corelist can error for both invalid core number but also for
incorrectly formatted input.
How about "invalid core list, please check core numbers are in [0, %u]
range" ?

diff --git a/test/test/test_eal_flags.c b/test/test/test_eal_flags.c
> index 2acab9d69..28e68a6c9 100644
> --- a/test/test/test_eal_flags.c
> +++ b/test/test/test_eal_flags.c
> @@ -513,6 +513,16 @@ test_missing_c_flag(void)
>         const char *argv25[] = { prgname, prefix, mp_flag,
>                                  "-n", "3", "--lcores",
>                                  "0-1,2@(5-7),(3-5)@(0,2),(0,6),7"};
> +       /* core number is negative value */
> +       const char * const argv26[] = { prgname, prefix, mp_flag,
> +                               "-n", "3", "--lcores", "-5" };
> +       const char * const argv27[] = { prgname, prefix, mp_flag,
> +                               "-n", "3", "--lcores", "-5-7" };
>
+       /* core number is maximum value */
> +       const char * const argv28[] = { prgname, prefix, mp_flag,
> +                               "-n", "3", "--lcores", "999999999" };
> +       const char * const argv29[] = { prgname, prefix, mp_flag,
> +                               "-n", "3", "--lcores", "1-9999999" };
>

Well, the maximum value is not really "999999999".
Please check against RTE_MAX_LCORE.


-- 
David Marchand