Skip to contentSkip to conversation

delay_tsc(): inefficient delay loop (2.6.16-mm1)

PATCH Andreas Mohr 3 participants 4 messages Mar 24, 2006

lore
Share & tools
lore
#0 lore raw

Hello all,

I discovered that the delay loop used there (which simply found a new place in 2.6.16-mm1; it existed much longer) is inefficient,
since it does an unnecessary subtraction *in the main loop* which also hits asm code:

00000024 <delay_tsc>:

  24:   53                      push   %ebx
  25:   89 c3                   mov    %eax,%ebx
  27:   0f 31                   rdtsc
  29:   89 c1                   mov    %eax,%ecx
  2b:   f3 90                   pause
  2d:   0f 31                   rdtsc
  2f:   29 c8                   sub    %ecx,%eax
  31:   39 d8                   cmp    %ebx,%eax
  33:   72 f6                   jb     2b <delay_tsc+0x7>
  35:   5b                      pop    %ebx
  36:   c3                      ret

With such a patch:

arch/i386/lib/delay.c renamed from arch/i386/lib/delay.c.orig +3 -1
@@ -44,10 +44,12 @@
 	unsigned long bclock, now;
 
 	rdtscl(bclock);
+	/* offset with bclock to have very simple comparison below */
+	loops += bclock;
 	do {
 		rep_nop();
 		rdtscl(now);
-	} while ((now-bclock) < loops);
+	} while (now < loops);
 }
 
 /*

the result is:

00000024 <delay_tsc>:

  24:   89 c1                   mov    %eax,%ecx
  26:   0f 31                   rdtsc
  28:   01 c1                   add    %eax,%ecx
  2a:   f3 90                   pause
  2c:   0f 31                   rdtsc
  2e:   39 c8                   cmp    %ecx,%eax
  30:   72 f8                   jb     2a <delay_tsc+0x6>
  32:   c3                      ret

Improvement: no unnecessary stuff after having hit the timer target value (read: faster), better power saving, 4 bytes less opcodes.

The patch above could be considered weird since it fiddles with "loops" directly. A possibly cleaner way would be to introduce a new variable end = bclock + loops.

Comments? Anything that I'm missing here as to why it hasn't been done that way before?

If this holds water then I'll submit a final patch soon.

Thanks!

Andreas Mohr

#1 lore raw
Quoted 1 line · Andreas Mohr

+ loops += bclock;

[...]

arch/i386/lib/delay.c line 52

Quoted 2 lines
arch/i386/lib/delay.c renamed from arch/i386/lib/delay.c.orig +1 -1
-       } while ((now-bclock) < loops);
+       } while (now < loops);

Erm, aren't you introducing an overflow problem here?

if loops is 2^32-1, bclock is 1, the old version would execute the proper number of times, the new one will blow out in one tick.

#2 lore raw

Hi,

Quoted 10 lines · Ray Lee
Quoted 1 line · Andreas Mohr

+ loops += bclock;

[...]

arch/i386/lib/delay.c line 52

Quoted 2 lines
arch/i386/lib/delay.c renamed from arch/i386/lib/delay.c.orig +1 -1
-       } while ((now-bclock) < loops);
+       } while (now < loops);

Erm, aren't you introducing an overflow problem here?

if loops is 2^32-1, bclock is 1, the old version would execute the proper number of times, the new one will blow out in one tick.

Doh. That's what happens if you get too excited about some new trick... Back to the drawing board, methinks.

Andreas Mohr

Signature · 2 lines
No programming skills!? Why not help translate many Linux applications! 
https://launchpad.ubuntu.com/rosetta
#3 lore raw
In reply to Andreas Mohr

arch/i386/lib/delay.c line 53

Quoted 10 lines · Andreas Mohr
arch/i386/lib/delay.c renamed from arch/i386/lib/delay.c.orig +3 -1
 	rdtscl(bclock);
+	/* offset with bclock to have very simple comparison below */
+	loops += bclock;
 	do {
 		rep_nop();
 		rdtscl(now);
-	} while ((now-bclock) < loops);
+	} while (now < loops);
 }

Hehe, optimizing delay loops *g* But your optimization is wrong. 'loops+bclock' and/or 'now' is likely to wrap around and then the test condition becomes bogus.

Ciao, ET.