From: Kalesh AP kalesh-anakkur.purayil@broadcom.com
[ Upstream commit 5bccacb4cc32cb835fe2fe100a210332c494e81d ]
The "ref_count" field in struct bnxt_ulp is unused after commit a43c26fa2e6c ("RDMA/bnxt_re: Remove the sriov config callback"). So we can just remove it now.
Reviewed-by: Somnath Kotur somnath.kotur@broadcom.com Signed-off-by: Kalesh AP kalesh-anakkur.purayil@broadcom.com Signed-off-by: Michael Chan michael.chan@broadcom.com Link: https://patch.msgid.link/20250417172448.1206107-4-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Sasha Levin sashal@kernel.org ---
**YES** This commit should be backported to stable kernel trees. **Analysis:** This commit removes an unused field `ref_count` from `struct bnxt_ulp` and the associated code that was waiting for the reference count to reach zero during unregistration. Looking at the code changes: 1. **Removes unused field from struct definition** (bnxt_ulp.h:49): The `atomic_t ref_count;` field is completely removed from `struct bnxt_ulp`. 2. **Removes dead code** (bnxt_ulp.c:151-162): The commit removes a loop that was waiting for `atomic_read(&ulp->ref_count) != 0` with sleeps, which is now unnecessary since the field and its usage were removed in a previous commit (a43c26fa2e6c). **Why this should be backported:** 1. **Code cleanup with no functional impact**: This is a pure cleanup commit that removes truly unused code. The `ref_count` field was made unused by a previous commit that removed the sriov config callback. 2. **Follows stable backport patterns**: Looking at the similar commits, particularly Similar Commit #1 which was marked "YES" for backport, this commit has the exact same characteristics: - Removes unused functions/fields - Small, contained change - No risk of regression - Code cleanup that improves maintainability 3. **No architectural changes**: The commit only removes code that was already dead/unused, with no changes to active code paths. 4. **Minimal risk**: Since the code being removed was already unused (as confirmed by the commit message referencing the previous commit that made it unused), there's zero risk of regression. 5. **Clear precedent**: Similar Commit #1 showed that removal of unused code (`bnxt_subtract_ulp_resources()` function and making `bnxt_get_max_func_irqs()` static) was considered appropriate for backporting. This commit fits the stable tree criteria perfectly: it's a low-risk cleanup that removes dead code without affecting functionality, similar to other cleanup commits that have been successfully backported.
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 5 ----- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 1 - 2 files changed, 6 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index e4a7f37036edb..40b2c477c239f 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -148,7 +148,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) struct net_device *dev = edev->net; struct bnxt *bp = netdev_priv(dev); struct bnxt_ulp *ulp; - int i = 0;
ulp = edev->ulp_tbl; rtnl_lock(); @@ -164,10 +163,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) synchronize_rcu(); ulp->max_async_event_id = 0; ulp->async_events_bmap = NULL; - while (atomic_read(&ulp->ref_count) != 0 && i < 10) { - msleep(100); - i++; - } mutex_unlock(&edev->en_dev_lock); rtnl_unlock(); return; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index 7fa3b8d1ebd28..f6b5efb5e7753 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -50,7 +50,6 @@ struct bnxt_ulp { unsigned long *async_events_bmap; u16 max_async_event_id; u16 msix_requested; - atomic_t ref_count; };
struct bnxt_en_dev {