Fix AST2700 eMMC kernel improvements and Arista Goldfinch DTS update

[prajjwal-arista]

Description

Add fix for AST2700 eMMC kernel improvements, and Arista Goldfinch DTS update for eMMC

Following is added in this patch to improve the eMMC -

1. Patch AspeedTech-BMC Linux Kernel "mmc: sdhci-of-aspeed: Optimize tuning mechanism" (v6.12) AspeedTech-BMC/linux@5fa88b1
2. Patch AspeedTech-BMC Linux Kernel "mmc: sdhci-of-aspeed: Improve CMD6 timing" (v6.18) AspeedTech-BMC/linux@e353214 (Note that this commit is not present in AspeedTech-BMC linux kernel branch v6.12, but in v6.18. We are patching it here in v6.12 as it is required for eMMC fix)
3. Update arista goldfinch dts to increase drive strength and use 8-bit data for eMMC

Why I did it

On few bmc units we faced eMMC issues resulting it in not being initialized properly during bootup-

$ sudo dmesg | grep -i mmc
[    4.499887] mmc0: SDHCI controller on 12090100.sdhci [12090100.sdhci] using ADMA 64-bit
[    4.720956] mmc0: mmc_select_hs200 failed, error -110
[    4.731344] mmc0: error -110 whilst initialising MMC card
[    5.039620] mmc0: mmc_select_hs200 failed, error -110
[    5.045306] mmc0: error -110 whilst initialising MMC card
[    5.328407] mmc0: mmc_select_hs200 failed, error -110
[    5.334089] mmc0: error -110 whilst initialising MMC card
[    5.665417] mmc0: mmc_select_hs200 failed, error -110
[    5.672118] mmc0: error -110 whilst initialising MMC card
[    5.683391] mmc0: Failed to initialize a non-removable card

Patch List

M patches-sonic/arista_goldfinch-dts.patch

• patches-sonic/mmc-sdhci-of-aspeed-Improve-CMD6-timing_v6.18.patch
• patches-sonic/mmc-sdhci-of-aspeed-Optimize-tuning-mechanism_v6.12.patch

How I did it

Added Aspeed patches for mmc sdhci for tuning and improving eMMC. Updating Arista Goldfinch BMC DTS to adapt the eMMC changes.

How i tested it

Built and booted up with the changes and saw no mmc related dmesg logs:

$ sudo dmesg | grep -i mmc
[    3.249624] mmc0: SDHCI controller on 12090100.sdhci [12090100.sdhci] using ADMA 64-bit
[    3.471941] mmc0: new HS200 MMC card at address 0001
[    3.480883] mmcblk0: mmc0:0001 TS0A32 9.71 GiB
[    3.494901]  mmcblk0: p1
[    3.500625] mmcblk0boot0: mmc0:0001 TS0A32 4.00 MiB
[    3.512853] mmcblk0boot1: mmc0:0001 TS0A32 4.00 MiB
[    3.522829] mmcblk0rpmb: mmc0:0001 TS0A32 4.00 MiB, chardev (238:0)

Backport

Backport required in 202605

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[vivekverma-arista]

@chander-nexthop @yxieca @judyjoseph

[yxieca]

Thanks for the focused fix and the verifiable dmesg evidence — the two kernel patches are byte-identical to the AspeedTech-BMC upstream commits, which made the verification easy.

One process suggestion before approval, on patch hygiene for the two cherry-picked kernel patches:

Both patches currently have From: Prajjwal Singh and drop the original Signed-off-by: Cool Lee <cool_lee@aspeedtech.com> chain. For patches cherry-picked verbatim from another tree, the kernel-community convention (see Documentation/process/submitting-patches.rst §11) is to preserve the original author + SOB, and add your own SOB plus a cherry-pick / upstream trailer.

Concrete suggestion: from a local clone of AspeedTech-BMC/linux, run

git format-patch -1 e3532140fbe86fe88abacde37aebbbd5520fbc9d
git format-patch -1 5fa88b13a96cf0059009f13cdf0b5b8b7b9c51ea

That gives you patch files that already carry the original From: and Signed-off-by: Cool Lee. Then amend each commit message with:

(cherry picked from commit <upstream-sha>)
Upstream: https://github.com/AspeedTech-BMC/linux/commit/<sha>
Signed-off-by: Prajjwal Singh <...>

This preserves git log/git blame attribution and makes it easy for future maintainers to drop the local patch after a kernel rebase by diffing against upstream.

Two smaller asks that would help reviewers:

1. Backport story — which stable branches does Goldfinch ship on? If any, this fix likely needs to go there too. Even a one-line "master only, Goldfinch not yet on a release branch" in the PR description clears the question.

2. Cross-board confirmation — the two kernel patches modify drivers/mmc/host/sdhci-of-aspeed.c, which is shared with the other AST2700 boards in this tree (nexthop b27, Nokia BMC H6-128). The new aspeed_sdhci_voltage_switch in particular now writes a register that was previously untouched. Could you confirm no regression on the other AST2700 boards (or tag those platform owners for sign-off)?

None of these are blockers on correctness — happy to approve once the patch headers are regenerated and the two questions above have a line of answer.

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[prajjwal-arista]

@yxieca Thank you for reviewing
I have addressed your comments, and also added the backport info (required in 202605).

Requesting @chander-nexthop @nats-nokia @oleksandrivantsiv to please review these changes.