[PATCH] nvme-multipath: require exact iopolicy names for module parameter
John Garry
john.g.garry at oracle.com
Mon Jun 1 08:25:26 PDT 2026
On 29/05/2026 09:51, dayou5941 at 163.com wrote:
> From: liyouhong <liyouhong at kylinos.cn>
>
> The iopolicy module parameter uses strncmp prefix matching, so values
> like "numax" are accepted as "numa". The per-subsystem sysfs attribute
> already requires an exact match via sysfs_streq(). Parse both through
> a shared helper so invalid values are rejected consistently.
>
> Signed-off-by: liyouhong <liyouhong at kylinos.cn>
Feel free to ignore my comments below, FWIW:
Reviewed-by: John Garry <john.g.garry at oracle.com>
> ---> drivers/nvme/host/multipath.c | 40
+++++++++++++++++++++--------------
> 1 file changed, 24 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
> index 263161cb8ac0..ce8dbb741d57 100644
> --- a/drivers/nvme/host/multipath.c
> +++ b/drivers/nvme/host/multipath.c
> @@ -73,19 +73,29 @@ static const char *nvme_iopolicy_names[] = {
>
> static int iopolicy = NVME_IOPOLICY_NUMA;
>
> +static int nvme_iopolicy_parse(const char *str)
> +{
> + int i;
> +
> + for (i = 0; i < ARRAY_SIZE(nvme_iopolicy_names); i++) {
> + if (sysfs_streq(str, nvme_iopolicy_names[i]))
> + return i;
> + }
> + return -EINVAL;
This is nearly duplicating __sysfs_match_string
BTW, going slightly off topic, I think that nvme_iopolicy_names could be
made const char *const
> +}
> +
> static int nvme_set_iopolicy(const char *val, const struct kernel_param *kp)
> {
> + int policy;
> +
> if (!val)
> return -EINVAL;
> - if (!strncmp(val, "numa", 4))
> - iopolicy = NVME_IOPOLICY_NUMA;
> - else if (!strncmp(val, "round-robin", 11))
> - iopolicy = NVME_IOPOLICY_RR;
> - else if (!strncmp(val, "queue-depth", 11))
> - iopolicy = NVME_IOPOLICY_QD;
> - else
> - return -EINVAL;
>
> + policy = nvme_iopolicy_parse(val);
> + if (policy < 0)
> + return policy;
> +
> + iopolicy = policy;
> return 0;
> }
>
> @@ -1039,16 +1049,14 @@ static ssize_t nvme_subsys_iopolicy_store(struct device *dev,
> {
> struct nvme_subsystem *subsys =
> container_of(dev, struct nvme_subsystem, dev);
> - int i;
> + int policy;
>
> - for (i = 0; i < ARRAY_SIZE(nvme_iopolicy_names); i++) {
> - if (sysfs_streq(buf, nvme_iopolicy_names[i])) {
> - nvme_subsys_iopolicy_update(subsys, i);
> - return count;
> - }
> - }
> + policy = nvme_iopolicy_parse(buf);
> + if (policy < 0)
> + return policy;
>
> - return -EINVAL;
> + nvme_subsys_iopolicy_update(subsys, policy);
> + return count;
> }
> SUBSYS_ATTR_RW(iopolicy, S_IRUGO | S_IWUSR,
> nvme_subsys_iopolicy_show, nvme_subsys_iopolicy_store);
More information about the Linux-nvme
mailing list