On Mon May 24, 2021 at 1:24 AM IDT, Daniil Stas wrote:
> Disabling clk_ck clock leads to link up status loss in phy, which
> leads to auto-negotiation restart before each network command
> execution.
>
> This issue is especially big for PXE boot protocol because of
> auto-negotiation restarts before each configuration filename trial.
>
> To avoid this issue don't disable clk_ck clock after it was enabled.
>
> Signed-off-by: Daniil Stas <daniil.s...@posteo.net>
> Cc: Ramon Fried <rfried....@gmail.com>
> Cc: Joe Hershberger <joe.hershber...@ni.com>
> Cc: Patrick Delaunay <patrick.delau...@foss.st.com>
> Cc: Patrice Chotard <patrice.chot...@foss.st.com>
> ---
> drivers/net/dwc_eth_qos.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/dwc_eth_qos.c b/drivers/net/dwc_eth_qos.c
> index e8242ca4e1..2f088c758f 100644
> --- a/drivers/net/dwc_eth_qos.c
> +++ b/drivers/net/dwc_eth_qos.c
> @@ -321,6 +321,7 @@ struct eqos_priv {
> void *rx_pkt;
> bool started;
> bool reg_access_ok;
> + bool clk_ck_enabled;
> };
>
> /*
> @@ -591,12 +592,13 @@ static int eqos_start_clks_stm32(struct udevice
> *dev)
> goto err_disable_clk_rx;
> }
>
> - if (clk_valid(&eqos->clk_ck)) {
> + if (clk_valid(&eqos->clk_ck) && !eqos->clk_ck_enabled) {
> ret = clk_enable(&eqos->clk_ck);
> if (ret < 0) {
> pr_err("clk_enable(clk_ck) failed: %d", ret);
> goto err_disable_clk_tx;
> }
> + eqos->clk_ck_enabled = true;
> }
> #endif
>
> @@ -648,8 +650,6 @@ static void eqos_stop_clks_stm32(struct udevice
> *dev)
> clk_disable(&eqos->clk_tx);
> clk_disable(&eqos->clk_rx);
> clk_disable(&eqos->clk_master_bus);
> - if (clk_valid(&eqos->clk_ck))
> - clk_disable(&eqos->clk_ck);
> #endif
>
> debug("%s: OK\n", __func__);
> --
> 2.31.0
Applied to u-boot-net/master, thanks!
Best regards,
Ramon Fried
