On Monday, 7 November 2016 11:11 Matthias Brugger wrote: > On 07/11/16 09:15, Andreas Färber wrote: > > Am 07.11.2016 um 09:06 schrieb Matthias Brugger: > >> On 6 de novembre de 2016 21:07:38 CET, "Andreas Färber" <[email protected]> > >> wrote: > >>> Kernel 4.9.rc3 from Kernel:HEAD is still busted due to symbol > >>> errors > >> > >> This should be fixed as soon as kernel:HEAD updates to v4.9-rc4 > > > > At least on armv7hl with Kernel:HEAD's 4.9.rc4 I still get: > > > > [ 5.209712] i2c_mv64xxx: no symbol version for memcpy > > [ 5.220935] i2c_mv64xxx: Unknown symbol memcpy (err -22) > > Ok thats something different, and looks really strange...
I've seen people reporting similar issues. The answer was that it's because of what's described in http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=84d69848c97f and that the solution is to disable CONFIG_MODVERSIONS. But that's apparently not what has been done in our master branch (and yet 4.9-rc* kernels from Kernel:HEAD work on x86_64). Michal Kubecek > > [ 5.226361] i2c_mv64xxx: no symbol version for __aeabi_idiv > > > > [ 5.260698] usbcore: no symbol version for _clear_bit > > > > Mounting Configuration File System... > > [ 5.269623] usbcore: Unknown symbol _clear_bit (err -22) > > [ 5.275382] libahci_platform: no symbol version for > > __gnu_mcount_nc [ 5.275385] libahci_platform: Unknown symbol > > __gnu_mcount_nc (err -22) [ 5.275433] i2c_mv64xxx: Unknown > > symbol __aeabi_idiv (err -22) [ 5.296496] i2c_mv64xxx: no symbol > > version for __gnu_mcount_nc [ 5.296499] i2c_mv64xxx: Unknown > > symbol __gnu_mcount_nc (err -22) [ OK [0[ 5.309807] usbcore: > > no symbol version for _test_and_set_bit m] Mounted Configuration > > File System. ... -- To unsubscribe, e-mail: [email protected] To contact the owner, e-mail: [email protected]
