[dpdk-dev] [PATCH v2 04/24] net/i40e/base: release spinlock before function returns

[Qi Zhang <qi.z.zhang@intel.com>]

Fix a bug that adminq spin lock is not be released before function
i40e_nvmupd_command returns.

Fixes: 5a2e46a3e405 ("net/i40e/base: refactor NVM update command processing")
Cc: stable@dpdk.org

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
---
 drivers/net/i40e/base/i40e_nvm.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c
index a1e78300d..946778ba3 100644
--- a/drivers/net/i40e/base/i40e_nvm.c
+++ b/drivers/net/i40e/base/i40e_nvm.c
@@ -907,7 +907,11 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw,
 
 	/* Acquire lock to prevent race condition where adminq_task
 	 * can execute after i40e_nvmupd_nvm_read/write but before state
-	 * variables (nvm_wait_opcode, nvm_release_on_done) are updated
+	 * variables (nvm_wait_opcode, nvm_release_on_done) are updated.
+	 *
+	 * During NVMUpdate, it is observed that lock could be held for
+	 * ~5ms for most commands. However lock is held for ~60ms for
+	 * NVMUPD_CSUM_LCB command.
 	 */
 	i40e_acquire_spinlock(&hw->aq.arq_spinlock);
 	switch (hw->nvmupd_state) {
@@ -930,7 +934,8 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw,
 		 */
 		if (cmd->offset == 0xffff) {
 			i40e_nvmupd_check_wait_event(hw, hw->nvm_wait_opcode);
-			return I40E_SUCCESS;
+			status = I40E_SUCCESS;
+			goto exit;
 		}
 
 		status = I40E_ERR_NOT_READY;
@@ -945,6 +950,7 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw,
 		*perrno = -ESRCH;
 		break;
 	}
+exit:
 	i40e_release_spinlock(&hw->aq.arq_spinlock);
 	return status;
 }
-- 
2.14.1