Messages in this thread | | | Date | Mon, 13 Jan 2020 06:54:06 -0800 | From | Jakub Kicinski <> | Subject | Re: [PATCH net-next v2 1/8] net: stmmac: Initial support for TBS |
| |
On Mon, 13 Jan 2020 14:02:36 +0100, Jose Abreu wrote: > Adds the initial hooks for TBS support. This needs a 32 byte descriptor > in order for it to work with current HW. Adds all the logic for Enhanced > Descriptors in main core but no HW related logic for now. > > Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h > index f98c5eefb382..dceaeb72a414 100644 > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h > @@ -42,10 +42,13 @@ struct stmmac_tx_info { > /* Frequently used values are kept adjacent for cache effect */ > struct stmmac_tx_queue { > u32 tx_count_frames; > + int tbs_avail; > + int tbs_en;
These could be bool or a bitfield?
> struct timer_list txtimer; > u32 queue_index; > struct stmmac_priv *priv_data; > struct dma_extended_desc *dma_etx ____cacheline_aligned_in_smp; > + struct dma_edesc *dma_entx ____cacheline_aligned_in_smp;
Won't this create a cache line-sized hole? Is the structure member supposed to be aligned or the data its pointing to?
> struct dma_desc *dma_tx; > struct sk_buff **tx_skbuff; > struct stmmac_tx_info *tx_skbuff_dma;
> diff --git a/include/linux/stmmac.h b/include/linux/stmmac.h > index 0531afa9b21e..19190c609282 100644 > --- a/include/linux/stmmac.h > +++ b/include/linux/stmmac.h > @@ -139,6 +139,7 @@ struct stmmac_txq_cfg { > u32 low_credit; > bool use_prio; > u32 prio; > + int tbs_en;
also bool?
> }; > > struct plat_stmmacenet_data {
| |