Messages in this thread | | | Date | Fri, 18 Jul 2008 00:17:03 +0400 | From | Alexey Dobriyan <> | Subject | Re: *sigh* /proc/*/pagemap |
| |
> > >> 5) check with microscope allocation there -- page-aligned address and > > size > > >> == 0 > > >> should allocate 0 bytes, and > > > > > > Where? > > > > kmalloc() in pagemap_read(). kmalloc(0) and integer wraparound look > > possible.
Now that I unsuccessfully tried to reproduce multiplication wraparound (on 32-bit), integer wraparound in kmalloc() simply can't happen here.
Relevant code from pagemap_read():
uaddr = (unsigned long)buf & PAGE_MASK; uend = (unsigned long)(buf + count); pagecount = (PAGE_ALIGN(uend) - uaddr) / PAGE_SIZE; pages = kmalloc(pagecount * sizeof(struct page *), GFP_KERNEL);
pagecount is effectively "count / PAGE_SIZE", where "count" is size_t. kmalloc() takes size_t as first argument, so one has to overflow
count / PAGE_SIZE * sizeof(pointer).
For sure, page is bigger than pointer.
Anyone still see holes?
| |