On Wed, 2 Sep 2026 09:56:49 +0800
[email protected] wrote:
> From: Jie Liu <[email protected]>
>
> This series updates the SXE2 poll mode driver (drivers/net/sxe2) and
> its common library (drivers/common/sxe2), fixing issues found in the
> previous version and re-aligning the driver with the reference
> implementation.
Very very close to being ready.
Two things left.
Patch 35/48: refresh link state on link change events
-------------------------------------------------------
Warning: the commit message says the call is removed from
sxe2_link_update_init(). It is not removed, it is made conditional:
ret = sxe2_drv_mac_link_status_get(adapter);
- if (ret) {
+ if (ret)
PMD_DEV_LOG_ERR(adapter, DRV, "Failed to get link status,
ret=%d", ret);
- goto l_end;
- }
-
- (void)sxe2_link_update(dev, 0);
+ else
+ (void)sxe2_link_update(dev, 0);
Keeping it is fine, the initial link state still has to come from
somewhere. Please just say that in the message instead of saying the
call is removed.
Patch 48/48: fix devargs parse error check for representors
-------------------------------------------------------------
The check itself is correct. rte_eth_devargs_parse() returns the
number of devargs parsed on success and a negative errno on error, so
"ret < 0" is right and the old test did break representor probing.
Warning: the error path below it still returns something else:
if (ret < 0) {
PMD_LOG_ERR(INIT, "Failed to parse device arguments:
%s",
dev->devargs->cls_str);
return -rte_errno;
}
rte_eth_devargs_parse() returns the error code directly, it does not
set rte_errno. If rte_errno happens to be 0 here, this returns 0,
and sxe2_eth_pmd_probe() tests "ret != 0" and carries on with an
empty eth_da. A parse failure then looks like a successful probe.
The line is not touched by the patch, but this patch is the one
fixing the error handling of that call, so please fix it here too:
return ret;
in both places.
