On 17/10/24 13:07, Peter Maydell wrote:
On Thu, 17 Oct 2024 at 16:29, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
By the way, it looks to me like QEMU has a regression
somewhere where we can't boot that sx1 test for the SD
card version -- it hangs during kernel boot waiting for
the MMC card. (An elderly QEMU binary I have boots OK.)
I'm looking into what's happened there.
Yes, you are correct. I did a quick check; the problem started with v9.1.
v9.0 boots fine.
It's an issue with commit 1ab08790bb75e4 -- when we did a refactor
of the SD card emulation we didn't notice that the omap mmc
controller was also using the sd_cmd_type_t enum and relied
on the values of that enum matching the meanings of the
different values of the controller's MMC_CMD register Type field.
I'm just testing a patch.
Doh indeed I missed that use in omap_mmc_command()...
Thanks for the quick analysis!
Phil.
