Commit 0991f070 authored by Geoff Thorpe's avatar Geoff Thorpe
Browse files

For whatever reason (compiler or header bugs), at least one commonly-used

linux system (namely mine) chokes on our definitions and uses of the "HZ"
symbol in crypto/tmdiff.[ch] and apps/speed.c as a "bad function cast"
(when in fact there is no function casting involved at all). In both cases,
it is easily worked around by not defining a cast into the macro and
jiggling the expressions slightly.

In addition - this highlights some cruft in openssl that needs sorting out.
The tmdiff.h header is exported as part of the openssl API despite the fact
that it is ugly as the driven sludge and not used anywhere in the library,
applications, or utilities. More weird still, almost identical code exists
in apps/speed.c though it looks to be slightly tweaked - so either tmdiff
should be updated and used by speed.c, or it should be dumped because it's
obviously not useful enough.

Rather than removing it for now, I've changed the API for tmdiff to at
least make sense. This involves taking the object type (MS_TM) from the
implementation and using it in the header rather than using "char *" in the
API and casting mercilessly in the code (ugh). If someone doesn't like
"MS_TM" and the "ms_time_***" naming, by all means change it. This should
be a harmless improvement, because the existing API is clearly not very
useful (eg. we reimplement it rather than using it in our own utils).

However, someone still needs to take a hack at consolidating speed.c and
tmdiff.[ch] somehow.
parent 2aaec9cc
Loading
Loading
Loading
Loading
+9 −0
Original line number Diff line number Diff line
@@ -4,6 +4,15 @@

 Changes between 0.9.7c and 0.9.8  [xx XXX xxxx]

  *) The tmdiff.h API was so ugly and minimal that our own timing utility
     (speed) prefers to use its own implementation. The two implementations
     haven't been consolidated as yet (volunteers?) but the tmdiff API has had
     its object type properly exposed (MS_TM) instead of casting to/from "char
     *". This may still change yet if someone realises MS_TM and "ms_time_***"
     aren't necessarily the greatest nomenclatures - but this is what was used
     internally to the implementation so I've used that for now.
     [Geoff Thorpe]

  *) Ensure that deprecated functions do not get compiled when
     OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
     the self-tests were still using deprecated key-generation functions so
+18 −4
Original line number Diff line number Diff line
@@ -208,11 +208,21 @@
#include <openssl/ecdh.h>
#endif

/*
 * The following "HZ" timing stuff should be sync'd up with the code in
 * crypto/tmdiff.[ch]. That appears to try to do the same job, though I think
 * this code is more up to date than libcrypto's so there may be features to
 * migrate over first. This is used in two places further down AFAICS. 
 * The point is that nothing in openssl actually *uses* that tmdiff stuff, so
 * either speed.c should be using it or it should go because it's obviously not
 * useful enough. Anyone want to do a janitorial job on this?
 */

/* The following if from times(3) man page.  It may need to be changed */
#ifndef HZ
# if defined(_SC_CLK_TCK) \
     && (!defined(OPENSSL_SYS_VMS) || __CTRL_VER >= 70000000)
#  define HZ ((double)sysconf(_SC_CLK_TCK))
#  define HZ sysconf(_SC_CLK_TCK)
# else
#  ifndef CLK_TCK
#   ifndef _BSD_CLK_TCK_ /* FreeBSD hack */
@@ -349,7 +359,8 @@ static double Time_F(int s)
		else
			{
			times(&tend);
			ret=((double)(tend.tms_utime-tstart.tms_utime))/HZ;
			ret = HZ;
			ret=(double)(tend.tms_utime-tstart.tms_utime) / ret;
			return((ret < 1e-3)?1e-3:ret);
			}
		}
@@ -2191,7 +2202,10 @@ show_res:
#endif
#ifdef HZ
#define as_string(s) (#s)
		printf("HZ=%g", (double)HZ);
		{
		double dbl = HZ;
		printf("HZ=%g", dbl);
		}
# ifdef _SC_CLK_TCK
		printf(" [sysconf value]");
# endif
+14 −15
Original line number Diff line number Diff line
@@ -106,7 +106,8 @@
#ifndef HZ
# if defined(_SC_CLK_TCK) \
     && (!defined(OPENSSL_SYS_VMS) || __CTRL_VER >= 70000000)
#  define HZ ((double)sysconf(_SC_CLK_TCK))
/* #  define HZ ((double)sysconf(_SC_CLK_TCK)) */
#  define HZ sysconf(_SC_CLK_TCK)
# else
#  ifndef CLK_TCK
#   ifndef _BSD_CLK_TCK_ /* FreeBSD hack */
@@ -120,7 +121,7 @@
# endif
#endif

typedef struct ms_tm
struct ms_tm
	{
#ifdef TIMES
	struct tms ms_tms;
@@ -136,9 +137,9 @@ typedef struct ms_tm
#    endif
#  endif
#endif
	} MS_TM;
	};

char *ms_time_new(void)
MS_TM *ms_time_new(void)
	{
	MS_TM *ret;

@@ -149,18 +150,17 @@ char *ms_time_new(void)
#ifdef OPENSSL_SYS_WIN32
	ret->thread_id=GetCurrentThread();
#endif
	return((char *)ret);
	return ret;
	}

void ms_time_free(char *a)
void ms_time_free(MS_TM *a)
	{
	if (a != NULL)
		OPENSSL_free(a);
	}

void ms_time_get(char *a)
void ms_time_get(MS_TM *tm)
	{
	MS_TM *tm=(MS_TM *)a;
#ifdef OPENSSL_SYS_WIN32
	FILETIME tmpa,tmpb,tmpc;
#endif
@@ -180,14 +180,13 @@ void ms_time_get(char *a)
#endif
	}

double ms_time_diff(char *ap, char *bp)
double ms_time_diff(MS_TM *a, MS_TM *b)
	{
	MS_TM *a=(MS_TM *)ap;
	MS_TM *b=(MS_TM *)bp;
	double ret;

#ifdef TIMES
	ret=(b->ms_tms.tms_utime-a->ms_tms.tms_utime)/HZ;
	ret = HZ;
	ret = (b->ms_tms.tms_utime-a->ms_tms.tms_utime) / ret;
#else
# ifdef OPENSSL_SYS_WIN32
	{
@@ -217,14 +216,14 @@ double ms_time_diff(char *ap, char *bp)
	return((ret < 0.0000001)?0.0000001:ret);
	}

int ms_time_cmp(char *ap, char *bp)
int ms_time_cmp(const MS_TM *a, const MS_TM *b)
	{
	MS_TM *a=(MS_TM *)ap,*b=(MS_TM *)bp;
	double d;
	int ret;

#ifdef TIMES
	d=(b->ms_tms.tms_utime-a->ms_tms.tms_utime)/HZ;
	d = HZ;
	d = (b->ms_tms.tms_utime-a->ms_tms.tms_utime) / d;
#else
# ifdef OPENSSL_SYS_WIN32
	d =(b->ms_win32.dwHighDateTime&0x000fffff)*10+b->ms_win32.dwLowDateTime/1e7;
+17 −5
Original line number Diff line number Diff line
@@ -59,6 +59,16 @@
/* Header for dynamic hash table routines
 * Author - Eric Young
 */
/* ... erm yeah, "dynamic hash tables" you say?
 * 
 * And what would dynamic hash tables have to do with any of this code *now*?
 * AFAICS, this code is only referenced by crypto/bn/exp.c which is an unused
 * file that I doubt compiles any more. speed.c is the only thing that could
 * use this (and it has nothing to do with hash tables), yet it instead has its
 * own duplication of all this stuff and looks, if anything, more complete. See
 * the corresponding note in apps/speed.c.
 * The Bemused - Geoff
 */

#ifndef HEADER_TMDIFF_H
#define HEADER_TMDIFF_H
@@ -67,11 +77,13 @@
extern "C" {
#endif

char *ms_time_new(void );
void ms_time_free(char *a);
void ms_time_get(char *a);
double ms_time_diff(char *start,char *end);
int ms_time_cmp(char *ap,char *bp);
typedef struct ms_tm MS_TM;

MS_TM *ms_time_new(void );
void ms_time_free(MS_TM *a);
void ms_time_get(MS_TM *a);
double ms_time_diff(MS_TM *start, MS_TM *end);
int ms_time_cmp(const MS_TM *ap, const MS_TM *bp);

#ifdef  __cplusplus
}