| Message ID | 20250317101551.1005706-2-m-malladi@ti.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | Bug fixes from XDP and perout series | expand |
On Mon, Mar 17, 2025 at 03:45:48PM +0530, Meghana Malladi wrote: > During network interface initialization, the NIC driver needs to register > its Rx queue with the XDP, to ensure the incoming XDP buffer carries a > pointer reference to this info and is stored inside xdp_rxq_info. > > While this struct isn't tied to XDP prog, if there are any changes in > Rx queue, the NIC driver needs to stop the Rx queue by unregistering > with XDP before purging and reallocating memory. Drop page_pool destroy > during Rx channel reset and this is already handled by XDP during > xdp_rxq_info_unreg (Rx queue unregister), failing to do will cause the > following warning: > > [ 271.494611] ------------[ cut here ]------------ > [ 271.494629] WARNING: CPU: 0 PID: 2453 at /net/core/page_pool.c:1108 0xffff8000808d5f60 I think it would be nice to include a bit more of the stack trace here. > > Fixes: 46eeb90f03e0 ("net: ti: icssg-prueth: Use page_pool API for RX buffer allocation") > Signed-off-by: Meghana Malladi <m-malladi@ti.com> It is a shame that we now have more asymmetry regarding the allocation of the pool and unwind on error prueth_prepare_rx_chan(). But if I see things correctly the freeing of the pool via xdp_rxq_info_unreg() is unconditional. And with that in mind I agree the approach taken by this patch makes sense. Reviewed-by: Simon Horman <horms@kernel.org> ...
Hi Simon, Thanks for reviewing the patch series. On 3/20/2025 6:23 PM, Simon Horman wrote: > On Mon, Mar 17, 2025 at 03:45:48PM +0530, Meghana Malladi wrote: >> During network interface initialization, the NIC driver needs to register >> its Rx queue with the XDP, to ensure the incoming XDP buffer carries a >> pointer reference to this info and is stored inside xdp_rxq_info. >> >> While this struct isn't tied to XDP prog, if there are any changes in >> Rx queue, the NIC driver needs to stop the Rx queue by unregistering >> with XDP before purging and reallocating memory. Drop page_pool destroy >> during Rx channel reset and this is already handled by XDP during >> xdp_rxq_info_unreg (Rx queue unregister), failing to do will cause the >> following warning: >> >> [ 271.494611] ------------[ cut here ]------------ >> [ 271.494629] WARNING: CPU: 0 PID: 2453 at /net/core/page_pool.c:1108 0xffff8000808d5f60 > > I think it would be nice to include a bit more of the stack trace here. > Sure, I will attach a link to the warning logs in v2. >> >> Fixes: 46eeb90f03e0 ("net: ti: icssg-prueth: Use page_pool API for RX buffer allocation") >> Signed-off-by: Meghana Malladi <m-malladi@ti.com> > > It is a shame that we now have more asymmetry regarding > the allocation of the pool and unwind on error prueth_prepare_rx_chan(). > > But if I see things correctly the freeing of the pool via > xdp_rxq_info_unreg() is unconditional. And with that in mind > I agree the approach taken by this patch makes sense. > Agreed on the asymmetry part, but I am not quite convinced on why xdp_rxq_info_unreg() is freeing the pool unconditionally, when the driver is the one which is allocating the pool. If xdp_rxq_info_unreg() is freeing it, then shouldn't xdp_rxq_info_reg() be allocating it ? > Reviewed-by: Simon Horman <horms@kernel.org> > > ...
diff --git a/drivers/net/ethernet/ti/icssg/icssg_common.c b/drivers/net/ethernet/ti/icssg/icssg_common.c index df5da7a98abf..afa01c22dee8 100644 --- a/drivers/net/ethernet/ti/icssg/icssg_common.c +++ b/drivers/net/ethernet/ti/icssg/icssg_common.c @@ -1216,9 +1216,6 @@ void prueth_reset_rx_chan(struct prueth_rx_chn *chn, prueth_rx_cleanup, !!i); if (disable) k3_udma_glue_disable_rx_chn(chn->rx_chn); - - page_pool_destroy(chn->pg_pool); - chn->pg_pool = NULL; } EXPORT_SYMBOL_GPL(prueth_reset_rx_chan);
During network interface initialization, the NIC driver needs to register its Rx queue with the XDP, to ensure the incoming XDP buffer carries a pointer reference to this info and is stored inside xdp_rxq_info. While this struct isn't tied to XDP prog, if there are any changes in Rx queue, the NIC driver needs to stop the Rx queue by unregistering with XDP before purging and reallocating memory. Drop page_pool destroy during Rx channel reset and this is already handled by XDP during xdp_rxq_info_unreg (Rx queue unregister), failing to do will cause the following warning: [ 271.494611] ------------[ cut here ]------------ [ 271.494629] WARNING: CPU: 0 PID: 2453 at /net/core/page_pool.c:1108 0xffff8000808d5f60 Fixes: 46eeb90f03e0 ("net: ti: icssg-prueth: Use page_pool API for RX buffer allocation") Signed-off-by: Meghana Malladi <m-malladi@ti.com> --- drivers/net/ethernet/ti/icssg/icssg_common.c | 3 --- 1 file changed, 3 deletions(-)