lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <D9XP2BH2CU02.1R9E1GSXQD9JB@cknow.org>
Date: Fri, 16 May 2025 17:36:03 +0200
From: "Diederik de Haas" <didi.debian@...ow.org>
To: "Algea Cao" <algea.cao@...k-chips.com>, <vkoul@...nel.org>,
 <kishon@...nel.org>, <heiko@...ech.de>, <cristian.ciocaltea@...labora.com>,
 <andy.yan@...k-chips.com>
Cc: <linux-phy@...ts.infradead.org>, <linux-arm-kernel@...ts.infradead.org>,
 <linux-rockchip@...ts.infradead.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] phy: phy-rockchip-samsung-hdptx: Fix PHY PLL output
 50.25MHz error

Hi,

On Sun Apr 27, 2025 at 11:51 AM CEST, Algea Cao wrote:
> When using HDMI PLL frequency division coefficient at 50.25MHz
> that is calculated by rk_hdptx_phy_clk_pll_calc(), it fails to
> get PHY LANE lock. Although the calculated values are within the
> allowable range of PHY PLL configuration.
>
> In order to fix the PHY LANE lock error and provide the expected
> 50.25MHz output, manually compute the required PHY PLL frequency
> division coefficient and add it to ropll_tmds_cfg configuration
> table.
>
> Signed-off-by: Algea Cao <algea.cao@...k-chips.com>
> Reviewed-by: Cristian Ciocaltea <cristian.ciocaltea@...labora.com>
>
> ---
>  drivers/phy/rockchip/phy-rockchip-samsung-hdptx.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/drivers/phy/rockchip/phy-rockchip-samsung-hdptx.c b/drivers/phy/rockchip/phy-rockchip-samsung-hdptx.c
> index fe7c05748356..77236f012a1f 100644
> --- a/drivers/phy/rockchip/phy-rockchip-samsung-hdptx.c
> +++ b/drivers/phy/rockchip/phy-rockchip-samsung-hdptx.c
> @@ -476,6 +476,8 @@ static const struct ropll_config ropll_tmds_cfg[] = {
>  	  1, 1, 0, 0x20, 0x0c, 1, 0x0e, 0, 0, },
>  	{ 650000, 162, 162, 1, 1, 11, 1, 1, 1, 1, 1, 1, 1, 54, 0, 16, 4, 1,
>  	  1, 1, 0, 0x20, 0x0c, 1, 0x0e, 0, 0, },
> +	{ 502500, 84, 84, 1, 1, 7, 1, 1, 1, 1, 1, 1, 1, 11, 1, 4, 5,
> +	  4, 11, 1, 0, 0x20, 0x0c, 1, 0x0e, 0, 0, },
>  	{ 337500, 0x70, 0x70, 1, 1, 0xf, 1, 1, 1, 1, 1, 1, 1, 0x2, 0, 0x01, 5,
>  	  1, 1, 1, 0, 0x20, 0x0c, 1, 0x0e, 0, 0, },
>  	{ 400000, 100, 100, 1, 1, 11, 1, 1, 0, 1, 0, 1, 1, 0x9, 0, 0x05, 0,

I found this patch in the 'fixes' branch in linux-phy:
https://git.kernel.org/pub/scm/linux/kernel/git/phy/linux-phy.git/commit/?h=fixes&id=f9475055b11c0c70979bd1667a76b2ebae638eb7

In the 'next' branch in linux-phy, I found this commit:
0edf9d2bb9b4 ("phy: rockchip: samsung-hdptx: Avoid Hz<->hHz unit conversion overhead")
https://git.kernel.org/pub/scm/linux/kernel/git/phy/linux-phy.git/commit/?h=next&id=0edf9d2bb9b4ba7566dfdc7605883e04575129d9

Where the values in ropll_tmds_cfg are converted from hHz to Hz and the
data type changes from u32 to unsigned long long.
But AFAICT it does NOT convert this '502500' value, which IIUC means
most values are in Hz, while this one is in hHz.

Am I missing something or should this new value also be converted?

Cheers,
  Diederik

Download attachment "signature.asc" of type "application/pgp-signature" (229 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ