Re^2: Insights sought into floating point division strangeness

perl was compiled with fastmath support

-ffast-math was one of the options from your -V log that caused me ask how your Perl was configured/built. The other was the -march=pentium2, which is probably perfectly valid, but seems way back level unless it's a very old machine or the -mcpu=athlon-xp requires it?

If your looking at your build configuration, it might be worth questioning these two also.

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

Lingua non convalesco, consenesco et abolesco. -- Rule 1 has a caveat! -- Who broke the cabal?

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

Comment on Re^2: Insights sought into floating point division strangeness Select or Download Code

Replies are listed 'Best First'.
Re^3: Insights sought into floating point division strangeness by abs0 (Beadle) on Feb 08, 2006 at 13:17 UTC
As the perpetrator of the aforementioned perl, we have a standard set of perl binaries across each os/arch, and as the lowest spec machine we have in service is a PII everything still gets -march=pentium2. The -ffast-math was probably a little over enthusiastic. Amusingly enough checking an alpha box shows it gets compiled with -mieee which ensure it has (slower) IEE 754 compliant behaviour, and so does not show this up. Needless to say our perl binaries are now !-ffast-math ...	[reply]
Re^4: Insights sought into floating point division strangeness by BrowserUk (Patriarch) on Feb 08, 2006 at 14:17 UTC
The defendant is accused of perpetrating a perl M'lud. A perl, {pause for effect} with fast-math no less. A most heinous crime for which, if found guilty, we will be demanding that you impose the most severe penalty possible. Does M'lud have his Black Cap to hand? When I first noticed the `pentium` in there, I wondered if the emulation was really that good :) I also spotted mention of A Pentium II bug, which also seemed to fit the description of the error, but given it's 1 in 8,589,934,592 odds of occurance, it seemed unlikely that it would occur at least 22 times as described in the OP. Anyway, I'm glad you sussed it out. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. Lingua non convalesco, consenesco et abolesco. -- Rule 1 has a caveat! -- Who broke the cabal? "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice.	[reply] [d/l]