mbox series

[0/6] msm8994-huawei-angler and msm8992-lg-bullhead sdhc1 fixes

Message ID 20220113233358.17972-1-petr.vorel@gmail.com
Headers show
Series msm8994-huawei-angler and msm8992-lg-bullhead sdhc1 fixes | expand

Message

Petr Vorel Jan. 13, 2022, 11:33 p.m. UTC
Hi,

Besides fixing sdhc1 for msm8994-huawei-angler and msm8992-lg-bullhead
I also rebased and reposted Jean's patches (Makefile didn't apply on
current for-next) and also rename angler's DTS.

sdhc1 (i.e. mmc0) is working on both, but there are overclocking
warnings. I suspected issue similar to using clk_rcg2_ops instead of
clk_rcg2_floor_ops (i.e. 3f905469c8ce ("clk: qcom: gcc: Use floor ops
for SDCC clocks") or 148ddaa89d4a ("clk: qcom: gcc-sc7180: Use floor ops
for the correct sdcc1 clk"), but it looks to be correct for gcc-msm8994.c.

Kind regards,
Petr

* angler
angler downstream
qcom,bus-speed-mode = "HS400_1p8v", "HS200_1p8v", "DDR_1p8v";
[90] SDHC Running in DDR mode
[    8.837072] sdhci: Secure Digital Host Controller Interface driver
[    8.843203] sdhci: Copyright(c) Pierre Ossman
[    8.847536] sdhci-pltfm: SDHCI platform and OF driver helper
[    8.858043] mmc0: no vqmmc regulator found
[    8.861168] mmc0: no vmmc regulator found
[    8.865359] mmc0: SDHCI controller on f9824900.sdhci [f9824900.sdhci] using 32-bit ADMA
[    8.920815] mmc1: no vqmmc regulator found
[    8.923917] mmc1: no vmmc regulator found
[    8.928123] mmc1: SDHCI controller on f98a4900.sdhci [f98a4900.sdhci] using 32-bit ADMA
[    9.018117] mmc0: BKOPS_EN bit = 0
[    9.039550] mmc0: new HS400 MMC card at address 0001
[    9.043971] mmcblk0: mmc0:0001 064G70 58.2 GiB
[    9.048185] mmcblk0rpmb: mmc0:0001 064G70 partition 3 4.00 MiB
[    9.056974] GPT:partition_entry_array_crc32 values don't match: 0x4aa75f3b != 0xe3345ddf
[    9.064044] GPT: Use GNU Parted to correct GPT errors.
[    9.069247]  mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 p41 p42 p43 p44

angler mainline (with this patchset)
using mmc-hs400-1_8v (this patchset)
[    1.825962] mmc0: Card appears overclocked; req 200000000 Hz, actual 384000000 Hz
[    1.826043] mmc0: Card appears overclocked; req 200000000 Hz, actual 384000000 Hz
[    1.839502] mmc0: Card appears overclocked; req 200000000 Hz, actual 384000000 Hz
[    1.840203] mmc0: Card appears overclocked; req 400000000 Hz, actual 768000000 Hz
Trying to mount subpartitions for 10 seconds...[    1.855276] mmc0: Card appears overclocked; req 400000000 Hz, actual 768000000 Hz
[    1.860454] mmc0: new HS400 MMC card at address 0001

[    1.878747] mmcblk0: mmc0:0001 064G70 58.2 GiB
[    1.883175] GPT:partition_entry_array_crc32 values don't match: 0x4aa75f3b != 0xe3345ddf
[    1.883234] GPT: Use GNU Parted to correct GPT errors.
[    1.894514]  mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 p41 p42 p43 p44
[    1.905939] mmcblk0boot0: mmc0:0001 064G70 4.00 MiB
[    1.926200] mmcblk0boot1: mmc0:0001 064G70 4.00 MiB
[    1.928880] mmcblk0rpmb: mmc0:0001 064G70 4.00 MiB, chardev (234:0)

using mmc-ddr-1_8v (and mmc-hs200-1_8v)
[    1.825991] mmc0: Card appears overclocked; req 200000000 Hz, actual 384000000 Hz
[    1.826071] mmc0: Card appears overclocked; req 200000000 Hz, actual 384000000 Hz
[    1.837905] mmc0: new HS200 MMC card at address 0001
[    1.846269] mmcblk0: mmc0:0001 064G70 58.2 GiB
[    1.850916] GPT:partition_entry_array_crc32 values don't match: 0x4aa75f3b != 0xe3345ddf
[    1.850976] GPT: Use GNU Parted to correct GPT errors.
Trying to mount subpartitions for 10 seconds...[    1.863540]  mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 p41 p42 p43 p44
[    1.879858] mmcblk0boot0: mmc0:0001 064G70 4.00 MiB

[    1.894920] mmcblk0boot1: mmc0:0001 064G70 4.00 MiB
[    1.899991] mmcblk0rpmb: mmc0:0001 064G70 4.00 MiB, chardev (234:0)

* bullhead 1.01
bullhead downstream
[70] SDHC Running in HS400 mode
[    7.069617] sdhci: Secure Digital Host Controller Interface driver
[    7.075741] sdhci: Copyright(c) Pierre Ossman
[    7.080081] sdhci-pltfm: SDHCI platform and OF driver helper
[    7.091419] mmc0: no vqmmc regulator found
[    7.094508] mmc0: no vmmc regulator found
[    7.098731] mmc0: SDHCI controller on f9824900.sdhci [f9824900.sdhci] using 32-bit ADMA
[    7.250588] mmc0: BKOPS_EN bit = 0
[    7.269765] mmc0: new HS400 MMC card at address 0001
[    7.274242] mmcblk0: mmc0:0001 032G74 29.1 GiB
[    7.278448] mmcblk0rpmb: mmc0:0001 032G74 partition 3 4.00 MiB
[    7.286719]  mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12 p13 p14 p15 p16 p17 p18 p19 p20 p21 p22 p23 p24 p25 p26 p27 p28 p29 p30 p31 p32 p33 p34 p35 p36 p37 p38 p39 p40 p41 p42 p43 p44 p45 p46

bullhead mainline (with this patchset)
[    1.961228] mmc0: Card appears overclocked; req 200000000 Hz, actual 343999999 Hz
[    1.961312] mmc0: Card appears overclocked; req 200000000 Hz, actual 343999999 Hz
[    1.968060] mmc0: Card appears overclocked; req 400000000 Hz, actual 687999999 Hz
Configuring kernel firmware image search path
/init: line 56: can't create /proc/sys/kernel/hotplug: nonexistent directory
[    1.988240] mmc0: Card appears overclocked; req 200000000 Hz, actual 343999999 Hz
[    1.998602] mmc0: Card appears overclocked; req 400000000 Hz, actual 687999999 Hz
[    2.001919] mmc0: Card appears overclocked; req 400000000 Hz, actual 687999999 Hz
[    2.018628] mmc0: new HS400 MMC card at address 0001

Jean THOMAS (2):
  arm64: dts: qcom: msm8992-lg-bullhead: Place LG Bullhead generic code
    into a DTSI file
  arm64: dts: qcom: msm8992-lg-bullhead: Add support for LG Bullhead rev
    1.0

Petr Vorel (4):
  arm64: dts: qcom: msm8994: Provide missing "xo_board" and "sleep_clk"
    to GCC
  arm64: dts: qcom: msm8994-huawei-angler: Add vendor name huawei
  arm64: dts: qcom: msm8994-huawei-angler: Add sdhc{1,2} definitions
  arm64: dts: qcom: msm8992-lg-bullhead: Disable sdhc2

 arch/arm64/boot/dts/qcom/Makefile                |  5 +++--
 .../boot/dts/qcom/msm8992-lg-bullhead-rev-10.dts | 14 ++++++++++++++
 .../dts/qcom/msm8992-lg-bullhead-rev-101.dts     | 14 ++++++++++++++
 ...head-rev-101.dts => msm8992-lg-bullhead.dtsi} |  7 +++++--
 ...101.dts => msm8994-huawei-angler-rev-101.dts} | 16 ++++++++++++++++
 arch/arm64/boot/dts/qcom/msm8994.dtsi            |  3 +++
 6 files changed, 55 insertions(+), 4 deletions(-)
 create mode 100644 arch/arm64/boot/dts/qcom/msm8992-lg-bullhead-rev-10.dts
 create mode 100644 arch/arm64/boot/dts/qcom/msm8992-lg-bullhead-rev-101.dts
 rename arch/arm64/boot/dts/qcom/{msm8992-bullhead-rev-101.dts => msm8992-lg-bullhead.dtsi} (98%)
 rename arch/arm64/boot/dts/qcom/{msm8994-angler-rev-101.dts => msm8994-huawei-angler-rev-101.dts} (75%)

Comments

patchwork-bot+linux-arm-msm@kernel.org Jan. 31, 2022, 9 p.m. UTC | #1
Hello:

This series was applied to qcom/linux.git (for-next)
by Bjorn Andersson <bjorn.andersson@linaro.org>:

On Fri, 14 Jan 2022 00:33:52 +0100 you wrote:
> Hi,
> 
> Besides fixing sdhc1 for msm8994-huawei-angler and msm8992-lg-bullhead
> I also rebased and reposted Jean's patches (Makefile didn't apply on
> current for-next) and also rename angler's DTS.
> 
> sdhc1 (i.e. mmc0) is working on both, but there are overclocking
> warnings. I suspected issue similar to using clk_rcg2_ops instead of
> clk_rcg2_floor_ops (i.e. 3f905469c8ce ("clk: qcom: gcc: Use floor ops
> for SDCC clocks") or 148ddaa89d4a ("clk: qcom: gcc-sc7180: Use floor ops
> for the correct sdcc1 clk"), but it looks to be correct for gcc-msm8994.c.
> 
> [...]

Here is the summary with links:
  - [1/6] arm64: dts: qcom: msm8992-lg-bullhead: Place LG Bullhead generic code into a DTSI file
    https://git.kernel.org/qcom/c/3f99518c6f65
  - [2/6] arm64: dts: qcom: msm8992-lg-bullhead: Add support for LG Bullhead rev 1.0
    https://git.kernel.org/qcom/c/cd4bd4704ec8
  - [3/6] arm64: dts: qcom: msm8994: Provide missing "xo_board" and "sleep_clk" to GCC
    (no matching commit)
  - [4/6] arm64: dts: qcom: msm8994-huawei-angler: Add vendor name huawei
    (no matching commit)
  - [5/6] arm64: dts: qcom: msm8994-huawei-angler: Add sdhc{1,2} definitions
    (no matching commit)
  - [6/6] arm64: dts: qcom: msm8992-lg-bullhead: Disable sdhc2
    (no matching commit)

You are awesome, thank you!
Bjorn Andersson Feb. 1, 2022, 5:19 a.m. UTC | #2
On Fri, 14 Jan 2022 00:33:52 +0100, Petr Vorel wrote:
> Besides fixing sdhc1 for msm8994-huawei-angler and msm8992-lg-bullhead
> I also rebased and reposted Jean's patches (Makefile didn't apply on
> current for-next) and also rename angler's DTS.
> 
> sdhc1 (i.e. mmc0) is working on both, but there are overclocking
> warnings. I suspected issue similar to using clk_rcg2_ops instead of
> clk_rcg2_floor_ops (i.e. 3f905469c8ce ("clk: qcom: gcc: Use floor ops
> for SDCC clocks") or 148ddaa89d4a ("clk: qcom: gcc-sc7180: Use floor ops
> for the correct sdcc1 clk"), but it looks to be correct for gcc-msm8994.c.
> 
> [...]

Applied, thanks!

[3/6] arm64: dts: qcom: msm8994: Provide missing "xo_board" and "sleep_clk" to GCC
      commit: 4dd1ad6192748523878463a285346db408b34a02
[4/6] arm64: dts: qcom: msm8994-huawei-angler: Add vendor name huawei
      commit: 8af90d6daa36a7180a2cd6aad874136aade27412

Best regards,