Re: Please test xen/stable-2.6.32.x in xen.git

["Olivier B." <xen.list@daevel.fr>]

On 31/03/2010 22:10, Olivier B. wrote:
> On 31/03/2010 19:58, Jeremy Fitzhardinge wrote:
>> Hi all,
>>
>> If you haven't already, could you please test out the 
>> xen/stable-2.6.32.x branch.  It has been working fairly well for me, 
>> but I'd like to get some wider testing coverage before we make it the 
>> default branch.
>>
>> To switch over, cd into your kernel directory, and type "git checkout 
>> xen/stable-2.6.32.x" and rebuild.
>>
>> This kernel is currently based on 2.6.32.10.  The main missing 
>> feature from xen/master is that it doesn't have netchannel2.  Is 
>> anybody using that yet?
>>
>> Thanks,
>>     J
>>
>
> Hi,
>
> for me it doesn't compile anymore :
>
>   CC      mm/vmalloc.o
> mm/vmalloc.c:582: error: expected declaration specifiers or ‘...’ 
> before ‘&’ token
> mm/vmalloc.c:582: warning: data definition has no type or storage class
> mm/vmalloc.c:582: warning: type defaults to ‘int’ in declaration of 
> ‘__spin_unlock’
> mm/vmalloc.c:582: warning: function declaration isn’t a prototype
> mm/vmalloc.c:582: error: conflicting types for ‘__spin_unlock’
> include/linux/spinlock_api_smp.h:349: note: previous definition of 
> ‘__spin_unlock’ was here
> mm/vmalloc.c:583: error: expected identifier or ‘(’ before ‘}’ token
> make[2]: *** [mm/vmalloc.o] Erreur 1
> make[1]: *** [mm] Erreur 2
> make[1]: quittant le répertoire « /home/daevel/linux/linux-xen-2.6.32.x »
> make: *** [debian/stamp/build/kernel] Erreur 2
>
> I compiled with "make-kpkg --revision=1.0 
> --append-to-version=-dae-dom0 kernel_image kernel_headers", and gcc 
> version 4.4.3 (Debian 4.4.3-3).
>
> Did I miss something ?
> Olivier

Same result with gcc version 4.4.3 (Debian 4.4.3-5)

I put here the code :

         if (nr)
                 atomic_sub(nr, &vmap_lazy_nr);

                 if (force_flush)
                         flush_tlb_kernel_range(*start, *end);

                 spin_lock(&vmap_area_lock);
                 list_for_each_entry_safe(va, n_va, &valist, purge_list)
                         __free_vmap_area(va);
                 spin_unlock(&vmap_area_lock);
         }
         spin_unlock(&purge_lock);

The ligne 582 in error is the "spin_unlock(&purge_lock);"