Messages in this thread | ![/](/images/icornerl.gif) | | Date | Tue, 27 Sep 2011 18:15:00 +0530 | From | Srikar Dronamraju <> | Subject | Re: [PATCH v5 3.1.0-rc4-tip 18/26] uprobes: slot allocation. |
| |
* Peter Zijlstra <peterz@infradead.org> [2011-09-27 14:18:52]:
> On Tue, 2011-09-20 at 17:33 +0530, Srikar Dronamraju wrote: > > +static struct uprobes_xol_area *xol_alloc_area(void) > > +{ > > + struct uprobes_xol_area *area = NULL; > > + > > + area = kzalloc(sizeof(*area), GFP_KERNEL); > > + if (unlikely(!area)) > > + return NULL; > > + > > + area->bitmap = kzalloc(BITS_TO_LONGS(UINSNS_PER_PAGE) * sizeof(long), > > + GFP_KERNEL); > > + > > + if (!area->bitmap) > > + goto fail; > > + > > + init_waitqueue_head(&area->wq); > > + spin_lock_init(&area->slot_lock); > > + if (!xol_add_vma(area) && !current->mm->uprobes_xol_area) { > > So what happens if xol_add_vma() succeeds, but we find > ->uprobes_xol_area set? > > > + task_lock(current); > > + if (!current->mm->uprobes_xol_area) { > > Having to re-test it under this lock seems to suggest it could. > > > + current->mm->uprobes_xol_area = area; > > + task_unlock(current); > > + return area; > > This function would be so much easier to read if the success case (this > here I presume) would not be nested 2 deep. > > > + } > > + task_unlock(current); > > + } > > at which point you could end up with two extra vmas? Because there's no > freeing of the result of xol_add_vma(). >
Agree, we need to unmap the vma in that case.
> > +fail: > > + kfree(area->bitmap); > > + kfree(area); > > + return current->mm->uprobes_xol_area; > > +}
-- Thanks and Regards Srikar
| ![\](/images/icornerr.gif) |