Hi Marcel,

Thanks for sharing the details.

1. I checked your log and noticed that the kernel cmdline already includes
   "clk_ignore_unused". Therefore, your issue does not appear to be the
   same as the one previously reported by Bo Gan.

2. I also verified that v6.19.2 already contains the two eMMC fix patches:
   https://lore.kernel.org/all/CAPDyKFqcMyM-=x+2FWNLhHY=gu5ApHNPQhp0xBKDJGr7BhEx4Q@mail.gmail.com/ (local)
   https://lore.kernel.org/all/CAPDyKFrkEZDuMbGpfxismcx=vJkSSK_XbtB762+sUFocupT63w@mail.gmail.com/ (local)

   A likely cause is that the eMMC DTS node is missing the AXI clock definition.
   Please refer to the DTS updates in the following pull request for reference:
   https://github.com/eswincomputing/linux-next/pull/20

3. In addition, please check whether CONFIG_GPIO_DWAPB=y is enabled in your
   kernel configuration. Since GMAC is built-in by default, it depends on
   GPIO being available during early initialization.

Hope this helps narrow things down. Please let me know what you find.

Best regards,
Zhi Li

-----原始邮件-----
发件人: "Marcel Ziswiler" [off-list ref]
发送时间:2026-02-21 02:37:57 (星期六)
收件人: lizhi2@eswincomputing.com, devicetree@vger.kernel.org, andrew+netdev@lunn.ch, davem@davemloft.net, edumazet@google.com, kuba@kernel.org, robh@kernel.org, krzk+dt@kernel.org, conor+dt@kernel.org, netdev@vger.kernel.org, pabeni@redhat.com, mcoquelin.stm32@gmail.com, alexandre.torgue@foss.st.com, rmk+kernel@armlinux.org.uk, linux-stm32@st-md-mailman.stormreply.com, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org
抄送: ningyu@eswincomputing.com, linmin@eswincomputing.com, pinkesh.vaghela@einfochips.com, weishangjuan@eswincomputing.com, dongxuyang@eswincomputing.com
主题: Re: [PATCH v2 0/2] net: stmmac: eic7700: fix EIC7700 eth1 RX sampling timing

Hi Zhi Li

Sorry, to steal this thread, but I do have a quick question concerning Ethernet on the EIC7700. Every time I
enable gmac0 in the device tree my EBC7700 stops booting (e.g. SD card is no longer detected) while Ethernet
does seem to at least get detected. This has been seen using next-20260219, stable v6.19.2 [1] as well as
6.18.0-rc6 [2]. Any idea what could be going wrong?

I admit so far I only do have one single EBC7700 at hand while waiting for further hardware to arrive.

Thanks!

[1] https://github.com/riscv/meta-riscv/pull/601
[2] https://github.com/eswincomputing/linux-next/tree/dev/test-upstream-v6.18-rc6

Cheers

Marcel

[snip]