kmemcheck
kmemcheck
Posted Nov 29, 2007 13:19 UTC (Thu) by dw (guest, #12017)Parent article: kmemcheck
Doubling the size of every allocation seems pretty wasteful, couldn't a bitmap be used instead, with byte or even word-level granularity? new_size = (size + (size/8))
Posted Nov 29, 2007 14:44 UTC (Thu)
by nix (subscriber, #2304)
[Link]
Posted Nov 29, 2007 15:20 UTC (Thu)
by jreiser (subscriber, #11027)
[Link]
Posted Nov 29, 2007 18:16 UTC (Thu)
by felixfix (subscriber, #242)
[Link] (3 responses)
Posted Nov 29, 2007 18:41 UTC (Thu)
by lysse (guest, #3190)
[Link] (2 responses)
Posted Nov 29, 2007 18:52 UTC (Thu)
by felixfix (subscriber, #242)
[Link]
Posted Nov 29, 2007 19:06 UTC (Thu)
by nix (subscriber, #2304)
[Link]
kmemcheck
A bitmap definitely can be used instead. (This is how valgrind's memcheck
tool implements uninitialized checks, although using JIT simulation rather
than page faulting to detect accesses.)
kmemcheck
On a 32-bit machine it is tempting to use 8 bits (an entire byte) to track access to 32 bits
(one word) of memory. This gives speed while still saving some space.
kmemcheck
The code to map bits would be slower; maybe he did not like the tradeoff.
kmemcheck
When there's already a page fault happening on every memory access, I have some doubts of the
significance of such additional overhead.
kmemcheck
Well, hmmmm, I guess I was thinking just enough to stick my foot in my mouth :-)
kmemcheck
Also the extra memory accesses will tend to blow the cache more if you use
more memory to track the uninit state.