Bitterblue Smith <rtl8821cerfe2@xxxxxxxxx> wrote:
> On 08/11/2024 21:55, Ezra Buehler wrote:
> > From: Ezra Buehler <ezra.buehler@xxxxxxxxxxxxxxxxxx>
> >
> > This has been tested on the AT91SAM9G25-based GARDENA smart Gateway
> > which uses an RTL8188CUS (product ID 0x8176).
> >
> > Signed-off-by: Ezra Buehler <ezra.buehler@xxxxxxxxxxxxxxxxxx>
> > ---
> >
> > Note that I was not able to figure out how many clients the RTL8192CU
> > and similar can actually handle, 127 is a wild guess.
> >
>
> I looked into that once and found that the old 8192cu and 8723au
> drivers have a limit of 32 macids.
>
> rtl8xxxu_fill_txdesc_v1() and rtl8xxxu_update_rate_mask() should
> fill the macid for AP mode to work well. Otherwise the firmware
> will not use the right TX rates for each connected client. It goes
> in the lowest 5 bits of tx_desc->txdw1 and h2c.ramask.arg.
>
> I tried AP mode in July 2023 and found that the TX speed was less
> than 10 megabits/second. In station mode it was at least 40.
> I didn't investigate more. How is the speed for you?
I also want to know the test result. Please mention that in commit
message as well.
>
> >
> > drivers/net/wireless/realtek/rtl8xxxu/8192c.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/drivers/net/wireless/realtek/rtl8xxxu/8192c.c
> b/drivers/net/wireless/realtek/rtl8xxxu/8192c.c
> > index 0abb1b092bc20..72918fe621fa1 100644
> > --- a/drivers/net/wireless/realtek/rtl8xxxu/8192c.c
> > +++ b/drivers/net/wireless/realtek/rtl8xxxu/8192c.c
> > @@ -644,6 +644,8 @@ struct rtl8xxxu_fileops rtl8192cu_fops = {
> > .rx_agg_buf_size = 16000,
> > .tx_desc_size = sizeof(struct rtl8xxxu_txdesc32),
> > .rx_desc_size = sizeof(struct rtl8xxxu_rxdesc16),
> > + .supports_ap = 1,
> > + .max_macid_num = 128,
As Bitterblue mentioned, this should be 32.
> > .max_sec_cam_num = 32,
> > .adda_1t_init = 0x0b1b25a0,
> > .adda_1t_path_on = 0x0bdb25a0,
> > --
> > 2.43.0
