RE: [PATCH v2] wifi: rtw88: usb: Further limit the TX aggregation

[Ping-Ke Shih <pkshih@realtek.com>]

Bitterblue Smith <rtl8821cerfe2@gmail.com> wrote:
> On 11/06/2024 06:26, Ping-Ke Shih wrote:
> > Bitterblue Smith <rtl8821cerfe2@gmail.com> wrote:
> >> @@ -2065,25 +2066,26 @@ const struct rtw_chip_info rtw8703b_hw_spec = {
> >>         .bt_rssi_type = COEX_BTRSSI_RATIO,
> >>         .ant_isolation = 15,
> >>         .rssi_tolerance = 2,
> >> -       .bt_rssi_step = bt_rssi_step_8703b,
> >> -       .wl_rssi_step = wl_rssi_step_8703b,
> >>         /* sant -> shared antenna, nsant -> non-shared antenna
> >>          * Not sure if 8703b versions with non-shard antenna even exist.
> >>          */
> >>         .table_sant_num = ARRAY_SIZE(table_sant_8703b),
> >> -       .table_sant = table_sant_8703b,
> >
> > Not sure why you move array and array size apart.
> > No need to churn this as well as other similar stuffs.
> >
> 
> I thought you wanted the members initialised in the same order
> they appear in struct rtw_chip_info in main.h. Maybe I misunderstood.

My bad, you didn't misunderstand. I didn't look into definition of
struct rtw_chip_info, which put the same data type together to save some space
because compiler can align fields:

	u8 table_sant_num;
	u8 table_nsant_num;
	u8 tdma_sant_num;
	u8 tdma_nsant_num;
	u8 afh_5g_num;
	u8 wl_rf_para_num;
	u8 coex_info_hw_regs_num;
	const u8 *bt_rssi_step;
	const u8 *wl_rssi_step;
	const struct coex_table_para *table_nsant;
	const struct coex_table_para *table_sant;
	const struct coex_tdma_para *tdma_sant;
	const struct coex_tdma_para *tdma_nsant;
	const struct coex_rf_para *wl_rf_para_tx;
	const struct coex_rf_para *wl_rf_para_rx;
	const struct coex_5g_afh_map *afh_5g;
	const struct rtw_hw_reg *btg_reg;
	const struct rtw_reg_domain *coex_info_hw_regs;

But in this table I would like to treat these as exceptions putting them together,
because each of them are related.
	.table_sant_num = ARRAY_SIZE(table_sant_8703b),
	.table_sant = table_sant_8703b,

Or if members of BT-coexistence part in tables across chips are in the same order,
just keep it as it was.

What do you think?

> 
> > This patch can only change things required by what subject mention. Align these
> > fields by another patch, or do it later. Both are fine to me.
> >
> 
> Ah, it should be done in a different patch, okay.

Yes. One patch is like v1, and the other one just changes the order, no logic change.