提交 c0366d89 编写于 作者: R Richard Levitte

3 changes:

- Make sure PCURSORINFO is defined even on systems that do not provide it.
- Change the reference to Peter Gutmann's paper.
- Make sure we don't walk the whole heap lists for performance reasons.
  Jeffrey Altman suggests following Peter Gutmann's advice to keep it
  to 50 heap entries per heap list.
上级 523d778a
......@@ -130,6 +130,15 @@
static void readtimer(void);
static void readscreen(void);
/* It appears like PCURSORINFO is only defined when WINVER is 0x0500 and up,
which currently only happens on Win2000. Unfortunately, that is a typedef,
so it's a little bit difficult to detect properly. On the other hand, the
macro CURSOR_SHOWING is defined within the same conditional, so it can be
use to detect the absence of PCURSORINFO. */
#ifndef CURSOR_SHOWING
typedef void *PCURSORINFO;
#endif
typedef BOOL (WINAPI *CRYPTACQUIRECONTEXT)(HCRYPTPROV *, LPCTSTR, LPCTSTR,
DWORD, DWORD);
typedef BOOL (WINAPI *CRYPTGENRANDOM)(HCRYPTPROV, DWORD, BYTE *);
......@@ -254,7 +263,7 @@ int RAND_poll(void)
*
* This seeding method was proposed in Peter Gutmann, Software
* Generation of Practically Strong Random Numbers,
* http://www.cs.auckland.ac.nz/~pgut001/pubs/random2.pdf
* http://www.usenix.org/publications/library/proceedings/sec98/gutmann.html
* (The assignment of entropy estimates below is arbitrary, but based
* on Peter's analysis the full poll appears to be safe. Additional
* interactive seeding is encouraged.)
......@@ -307,10 +316,14 @@ int RAND_poll(void)
if (heap_first(&hentry,
hlist.th32ProcessID,
hlist.th32HeapID))
{
int entrycnt = 50;
do
RAND_add(&hentry,
hentry.dwSize, 0);
while (heap_next(&hentry));
while (heap_next(&hentry)
&& --entrycnt > 0);
}
} while (heaplist_next(handle,
&hlist));
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册