On Wed, 2016-10-19 at 18:54 +0200, Sebastian Andrzej Siewior wrote:
> On 2016-10-19 17:56:30 [+0200], Mike Galbraith wrote:
> > In v4.7, the driver switched to percpu compression streams, disabling
> > preemption vai get/put_cpu_ptr(). Use a local lock instead for RT.
> > We also have to fix an RT lock order issue in zram_decompress_page()
> > such that zs_map_object() nests inside of zcomp_stream_put() as it
> > does in zram_bvec_write().
>
> good. I almost had it myself. So let me get that one. And your previous
> one (the spinlock replacement) looks also reasonable. With this hunk
>
> @@ -313,6 +314,7 @@ struct mapping_area {
> #endif
> char *vm_addr; /* address of kmap_atomic()'ed pages */
> enum zs_mapmode vm_mm; /* mapping mode */
> + spinlock_t ma_lock;
> };
>
> #ifdef CONFIG_COMPACTION
> @@ -1489,6 +1491,7 @@ void *zs_map_object(struct zs_pool *pool, unsigned long handle,
> off = (class->size * obj_idx) & ~PAGE_MASK;
>
> area = per_cpu_ptr(&zs_map_area, get_cpu_light());
> + spin_lock(&area->ma_lock);
> area->vm_mm = mm;
> if (off + class->size <= PAGE_SIZE) {
> /* this object is contained entirely within a page */
> @@ -1542,6 +1545,7 @@ void zs_unmap_object(struct zs_pool *pool, unsigned long handle)
>
> __zs_unmap_object(area, pages, off, class->size);
> }
> + spin_unlock(&area->ma_lock);
> put_cpu_light();
>
> migrate_read_unlock(zspage);
Ew, yeah, as the other, I thought it was covered, but nope.
-Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
