]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
i40e: remove redundant memory barrier when cleaning Tx descs
authorMaciej Fijalkowski <maciej.fijalkowski@intel.com>
Fri, 22 Aug 2025 15:16:17 +0000 (17:16 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 2 Oct 2025 11:35:41 +0000 (13:35 +0200)
commite69eb3d97a0ac1fc64c007122914c34148fca6d1
tree17e81375c61f4f7cdac6df191cdc28d97dfa4a64
parent35c78a7643461478e42e7670efabb6d3f364920a
i40e: remove redundant memory barrier when cleaning Tx descs

[ Upstream commit e37084a26070c546ae7961ee135bbfb15fbe13fd ]

i40e has a feature which writes to memory location last descriptor
successfully sent. Memory barrier in i40e_clean_tx_irq() was used to
avoid forward-reading descriptor fields in case DD bit was not set.
Having mentioned feature in place implies that such situation will not
happen as we know in advance how many descriptors HW has dealt with.

Besides, this barrier placement was wrong. Idea is to have this
protection *after* reading DD bit from HW descriptor, not before.
Digging through git history showed me that indeed barrier was before DD
bit check, anyways the commit introducing i40e_get_head() should have
wiped it out altogether.

Also, there was one commit doing s/read_barrier_depends/smp_rmb when get
head feature was already in place, but it was only theoretical based on
ixgbe experiences, which is different in these terms as that driver has
to read DD bit from HW descriptor.

Fixes: 1943d8ba9507 ("i40e/i40evf: enable hardware feature head write back")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Rinitha S <sx.rinitha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/net/ethernet/intel/i40e/i40e_txrx.c