Re: [PATCH] habanalabs: select CRC32

[Vegard Nossum <vegard.nossum@oracle.com>]

On 10/11/21 9:37 PM, Arnd Bergmann wrote:
> On Mon, Oct 11, 2021 at 6:29 PM Vegard Nossum <vegard.nossum@oracle.com> wrote:
>> I also suspect that satrandconfig samples the configuration space MUCH
>> more uniformly than randconfig, in the sense that "make randconfig" will
>> tend to have CRC32 enabled just because there are so many drivers that
>> _do_ select it. (Think of it this way, every prompt for an option that
>> has "select CRC32" in it will be another 50% probability of actually
>> enabling the option.)
> 
> Yes, of course. I've seen other problems like this often happen with
> 
> CONFIG_I2C=m
> CONFIG_LEDS_CLASS=m
> CONFIG_IPV6=m
> CONFIG_CRYPTO=m
> CONFIG_PM_SLEEP=n
> 
> Not sure if you're already hitting those or if you need to add something
> to your setup to make those more easy to find.

The CRYPTO one I've definitely hit a few times (usually combined with
CRYPTO_HASH) and a failure to link crypto_ahash_* functions.

Otherwise it's mostly CRC32 and a couple of odd ones like this that I've
seen so far:

ERROR: modpost: "dell_privacy_has_mic_mute"
[drivers/platform/x86/dell/dell-laptop.ko] undefined!

  CC      drivers/scsi/aic7xxx/aic79xx_reg_print.o
gcc: error: drivers/scsi/aic7xxx/aic79xx_reg_print.c: No such file or
directory
gcc: fatal error: no input files
compilation terminated.
make[3]: *** [scripts/Makefile.build:271:
drivers/scsi/aic7xxx/aic79xx_reg_print.o] Error 1

I'll give your suggestions a try, I have about 9 patches queued up so
far so I think there are definitely more problems lurking around.

Thanks,


Vegard