Message ID | 20231122155117.2816724-1-ioana.ciornei@nxp.com (mailing list archive) |
---|---|
Headers | show |
Series | dpaa2-eth: various fixes | expand |
On Wed, Nov 22, 2023 at 05:51:14PM +0200, Ioana Ciornei wrote: > The first two patches fix a memory corruption issue happening between > the Tx and Tx confirmation of a packet by making the Tx alignment at > 64bytes mandatory instead of optional as it was previously. > > The third patch fixes the Rx copybreak code path which recycled the > initial data buffer before all processing was done on the packet. > I think patches 1&2 should form a single patch, because they are supposed to be backported to older stable kernels and this is hard to do, if one of patches lacks "Fixes" tag. At the same time, they clearly complement each other. > Ioana Ciornei (3): > dpaa2-eth: increase the needed headroom to account for alignment > dpaa2-eth: set needed_headroom for net_device > dpaa2-eth: recycle the RX buffer only after all processing done > > drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 16 ++++++++++------ > drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h | 2 +- > 2 files changed, 11 insertions(+), 7 deletions(-) > > -- > 2.25.1 > >
On Wed, Nov 22, 2023 at 07:56:12PM +0100, Larysa Zaremba wrote: > On Wed, Nov 22, 2023 at 05:51:14PM +0200, Ioana Ciornei wrote: > > The first two patches fix a memory corruption issue happening between > > the Tx and Tx confirmation of a packet by making the Tx alignment at > > 64bytes mandatory instead of optional as it was previously. > > > > The third patch fixes the Rx copybreak code path which recycled the > > initial data buffer before all processing was done on the packet. > > > > I think patches 1&2 should form a single patch, because they are supposed to be > backported to older stable kernels and this is hard to do, if one of patches > lacks "Fixes" tag. At the same time, they clearly complement each other. Ok, I will squash the two patches and submit again.