Ensure we mask freelist pointers properly on 32-bit Linux

Ensure we mask freelist pointers properly on 32-bit Linux

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=158485

[jschuh, 2012-09-24 20:32:40]

My first patch could hit an odd corner-case when a 32-bit Linux process is
running on a 64-bit system. This patch corrects it by guaranteeing the masking
operation always results in a non-zero pointer. I also changed the typecast used
in the bit-shift, to improve the odds that the mask will point to an invalid
section of memory in this scenario.

[jar (doing other things), 2012-09-24 21:59:54]

lgtm

https://codereview.chromium.org/10957067/diff/2001/third_party/tcmalloc/chrom...
File third_party/tcmalloc/chromium/src/free_list.cc (right):

https://codereview.chromium.org/10957067/diff/2001/third_party/tcmalloc/chrom...
third_party/tcmalloc/chromium/src/free_list.cc:85: const uintptr_t q =
~(reinterpret_cast<uintptr_t>(TCMalloc_SystemAlloc)
nit: rename q--> mask

[commit-bot: I haz the power, 2012-09-24 22:05:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jschuh@chromium.org/10957067/6001

[jschuh, 2012-09-24 22:06:00]

https://codereview.chromium.org/10957067/diff/2001/third_party/tcmalloc/chrom...
File third_party/tcmalloc/chromium/src/free_list.cc (right):

https://codereview.chromium.org/10957067/diff/2001/third_party/tcmalloc/chrom...
third_party/tcmalloc/chromium/src/free_list.cc:85: const uintptr_t q =
~(reinterpret_cast<uintptr_t>(TCMalloc_SystemAlloc)
On 2012/09/24 21:59:54, jar wrote:
> nit: rename q--> mask

Done.

[commit-bot: I haz the power, 2012-09-25 01:10:58]

Change committed as 158485