| 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
| ||
|
Message-ID: <8c48de90-3dda-40c8-8f7a-01102f7360a3@quicinc.com>
Date: Fri, 21 Feb 2025 18:43:45 -0800
From: Wesley Cheng <quic_wcheng@...cinc.com>
To: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
CC: Melody Olvera <quic_molvera@...cinc.com>, Vinod Koul <vkoul@...nel.org>,
Kishon Vijay Abraham I <kishon@...nel.org>,
Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley <conor+dt@...nel.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Philipp Zabel
<p.zabel@...gutronix.de>,
Catalin Marinas <catalin.marinas@....com>,
"Will
Deacon" <will@...nel.org>,
Bjorn Andersson <andersson@...nel.org>,
"Konrad
Dybcio" <konradybcio@...nel.org>,
Satya Durga Srinivasu Prabhala
<quic_satyap@...cinc.com>,
Trilok Soni <quic_tsoni@...cinc.com>, <linux-arm-msm@...r.kernel.org>,
<linux-phy@...ts.infradead.org>, <devicetree@...r.kernel.org>,
<linux-kernel@...r.kernel.org>, <linux-usb@...r.kernel.org>,
<linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH 4/7] phy: qcom: qmp-combo: Add new PHY sequences for
SM8750
On 2/4/2025 6:59 AM, Dmitry Baryshkov wrote:
> On Mon, Feb 03, 2025 at 07:31:29PM -0800, Wesley Cheng wrote:
>>
>> On 1/14/2025 2:23 AM, Dmitry Baryshkov wrote:
>>> On Mon, Jan 13, 2025 at 01:52:10PM -0800, Melody Olvera wrote:
>>>> From: Wesley Cheng <quic_wcheng@...cinc.com>
>>>>
>>>> Add new register offsets and PHY values for SM8750. Some of the previous
>>>> definitions can be leveraged from older PHY versions as offsets within
>>>> registers have not changed. This also updates the PHY sequence that is
>>>> recommended after running hardware characterization.
>>>>
>>>> Signed-off-by: Wesley Cheng <quic_wcheng@...cinc.com>
>>>> Signed-off-by: Melody Olvera <quic_molvera@...cinc.com>
>>>> ---
>>>> drivers/phy/qualcomm/phy-qcom-qmp-combo.c | 198 ++++++++++++++++++++++++++++++
>>>> 1 file changed, 198 insertions(+)
>>>>
>>>> diff --git a/drivers/phy/qualcomm/phy-qcom-qmp-combo.c b/drivers/phy/qualcomm/phy-qcom-qmp-combo.c
>>>> index b09fa00e9fe7db8d97b7179ee15d3f07fe578b0c..823a60029ea6acbd1f0f8c7d27aaa58de39ed758 100644
>>>> --- a/drivers/phy/qualcomm/phy-qcom-qmp-combo.c
>>>> +++ b/drivers/phy/qualcomm/phy-qcom-qmp-combo.c
>>>> @@ -1471,6 +1471,139 @@ static const struct qmp_phy_init_tbl x1e80100_usb43dp_pcs_tbl[] = {
>>>> QMP_PHY_INIT_CFG(QPHY_V6_N4_PCS_EQ_CONFIG5, 0x10),
>>>> };
>>>>
>>>> +static const struct qmp_phy_init_tbl sm8750_usb3_serdes_tbl[] = {
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_STEP_SIZE1_MODE1, 0xc0),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_STEP_SIZE2_MODE1, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_CP_CTRL_MODE1, 0x02),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_PLL_RCTRL_MODE1, 0x16),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_PLL_CCTRL_MODE1, 0x36),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_CORECLK_DIV_MODE1, 0x04),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_LOCK_CMP1_MODE1, 0x16),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_LOCK_CMP2_MODE1, 0x41),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DEC_START_MODE1, 0x41),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DEC_START_MSB_MODE1, 0x00),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START1_MODE1, 0x55),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START2_MODE1, 0x75),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START3_MODE1, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_HSCLK_SEL_1, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_VCO_TUNE1_MODE1, 0x25),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_VCO_TUNE2_MODE1, 0x02),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_BIN_VCOCAL_CMP_CODE1_MODE1, 0x5c),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_BIN_VCOCAL_CMP_CODE2_MODE1, 0x0f),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_BIN_VCOCAL_CMP_CODE1_MODE0, 0x5c),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_BIN_VCOCAL_CMP_CODE2_MODE0, 0x0f),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_STEP_SIZE1_MODE0, 0xc0),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_STEP_SIZE2_MODE0, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_CP_CTRL_MODE0, 0x02),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_PLL_RCTRL_MODE0, 0x16),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_PLL_CCTRL_MODE0, 0x36),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_LOCK_CMP1_MODE0, 0x08),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_LOCK_CMP2_MODE0, 0x1a),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DEC_START_MODE0, 0x41),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DEC_START_MSB_MODE0, 0x00),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START1_MODE0, 0x55),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START2_MODE0, 0x75),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_DIV_FRAC_START3_MODE0, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_VCO_TUNE1_MODE0, 0x25),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_VCO_TUNE2_MODE0, 0x02),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_BG_TIMER, 0x0a),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_EN_CENTER, 0x01),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_PER1, 0x62),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SSC_PER2, 0x02),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SYSCLK_BUF_ENABLE, 0x0c),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_SYSCLK_EN_SEL, 0x1a),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_LOCK_CMP_CFG, 0x14),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_VCO_TUNE_MAP, 0x04),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_CORE_CLK_EN, 0x20),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_CMN_CONFIG_1, 0x16),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_AUTO_GAIN_ADJ_CTRL_1, 0xb6),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_AUTO_GAIN_ADJ_CTRL_2, 0x4a),
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_AUTO_GAIN_ADJ_CTRL_3, 0x36),
>>> The only difference from sm8550_usb3_serdes_tbl, it has 0x37 here.
>> Not sure what the suggestion is here. I think in general I would want to have a separate table for each chipset, considering that settings may change/evolve. Currently, if you're asking to re-use the sm8550 table to avoid re-defining this sequence, I think it'll be confusing to folks when they refer to this SOC's PHY settings.
>
>
> No suggestion, merely a question if both tables are correct or not.
>
>>>> + QMP_PHY_INIT_CFG(QSERDES_V6_COM_ADDITIONAL_MISC, 0x0c),
>>>> +};
>>>> +
>>> [...]
>>>
>>>> @@ -1781,6 +1914,22 @@ static const struct qmp_combo_offsets qmp_combo_offsets_v5 = {
>>>> .dp_dp_phy = 0x2200,
>>>> };
>>>>
>>>> +static const struct qmp_combo_offsets qmp_combo_offsets_v8 = {
>>> Why is it v8? Is the actual PHY also v8 rather than v6?
>> Yes, actual QMP PHY major rev is v8. If we want to, I can generate a separate v8 based header files if that is better. However, most of the offsets for the registers we're taking advantage of in the actual driver has the same offsets as previous revisions.
>
> If all registers of a particular set (QSERDE, TX/RX, PCS) are the same,
> then it should be fine to reuse those (although it creates some
> questions). If the majority is the same, but there are some differences,
> please create new header file. It is definitely easier to verify that
> the patch is correct if every piece has the same version.
>
Hi Dmitry,
I reviewed the differences in the register sets, and there are new addtions
in the v8 QMP PHY, however, we do not use or program those registers as of
now. I noticed as well in previous versions, we only capture register
defines if they are used in any of the init sequences. Assuming this
applies in this scenario, its probably ok to keep the v6 version until we
actually need to add any settings for the new registers.
Thanks
Wesley Cheng
Powered by blists - more mailing lists