Re: [PATCH 09/23] hw/sd/sdcard: Generate random RCA value

["Philippe Mathieu-Daudé" <philmd@linaro.org>]

On 21/6/24 11:44, Philippe Mathieu-Daudé wrote:
> On 21/6/24 10:05, Philippe Mathieu-Daudé wrote:
>> Rather than using the obscure 0x4567 magic value,
>> use a real random one.
>>
>> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
>> ---
>>   hw/sd/sd.c | 8 ++------
>>   1 file changed, 2 insertions(+), 6 deletions(-)


> Hmm the NPCM7xx test is failing:
> 
> ▶  58/450 ERROR:../tests/qtest/npcm7xx_sdhci-test.c:101:sdwrite_read: 
> assertion failed: (!memcmp(rmsg, msg, len)) ERROR
> 
> But booting various Linux / FreeBSD guests on SD cards works,
> so I suppose the test (or TYPE_NPCM7XX_SDHCI model) need some
> rework.
> 
> $ git grep 0x4567
> tests/qtest/npcm7xx_sdhci-test.c:47:    sdhci_cmd_regs(qts, 
> NPCM7XX_MMC_BA, 0, 0, 0x45670000, 0,
> tests/qtest/npcm7xx_sdhci-test.c-48- SDHC_SELECT_DESELECT_CARD);
> 
> In tests/qtest/libqos/sdhci-cmd.h:
> 
> /* CMD Reg */
> #define SDHC_SEND_RELATIVE_ADDR (3 << 8)
> #define SDHC_SELECT_DESELECT_CARD (7 << 8)
> 
> IIUC setup_sd_card():
> 
> card address is queried ...:
> 
>      sdhci_cmd_regs(qts, NPCM7XX_MMC_BA, 0, 0, 0, 0,
>                     SDHC_SEND_RELATIVE_ADDR);
> 
> ... but then ignored and magic 0x4567 is used instead:
> 
>      sdhci_cmd_regs(qts, NPCM7XX_MMC_BA, 0, 0, 0x45670000, 0,
>                     SDHC_SELECT_DESELECT_CARD);
> 
> OK, so this test need rework. I see the sdhci_read_cmd()
> method but it reads the SDHC_BDATA FIFO register, not sure
> this is what should be used to read the RCA from R6 command
> response.
> 
> Shengtan, Nuvoton folks, what do you think?

More than 1 week passed so I'll have a look at it myself.

Regards,

Phil.