[Openchrome-users] Driver fails to load after update

[Thomas Hellström]

Xavier Bachelot wrote:

>> Could you do a nm on all .o files and see which one contains the 
>> offending symbol? I don't have it here.
>>
> hmm, not sure which one is the offending symbol, so here the grep for 
> both :
>
> [xavierb at kingston via]$ grep XAACopyROP *.o
> Binary file via_accel.o matches
> Binary file via_drv.o matches
> [xavierb at kingston via]$ grep vgaHWGetIndex *.o
> Binary file via_bandwidth.o matches
> Binary file via_ch7xxx.o matches
> Binary file via_driver.o matches
> Binary file via_drv.o matches
> Binary file via_i2c.o matches
> Binary file via_id.o matches
> Binary file via_mode.o matches
> Binary file via_swov.o matches
> Binary file via_vbe.o matches
> Binary file via_vt162x.o matches
>
> This is from the build where I added vgaHWGetIndex to the symbol list.
>
> Xavier

Hi.
vgaHWGetIndex should be in the symbol list and has been added in SVN. 
It's referenced by a macro.

The situation is different with XAACopyROP and I wonder what makes your 
build different. Other people are successfully using latest SVN.

I can't  track it down without being able to reproduce, but what would 
help would be if you looked at the compile command for via_accel.c, i.e.

cc <a LOT of flags > -o via_accel.o -c via_accel.c

and manually did

cc <the same flags> -E via_accel.c > preprocessed.c 

That will cause the preprocessor only to run and output a preprocessed file.
If you could mail me that file together with the actual "cc" command 
used it should be possible to see what's happening.

/Thomas