AMD k6 in sid/sarge.
I have an ?old? AMD k6(3d). My mother uses this PC for surfing(printing web pages) and e-mail. I
have recently had problems with the new mozilla, bug #226358.
> I realy do HATE it when gcc makes asm like this. Whatever can loop will loop.
> 0x080714f0 <_ZN12nsStrPrivate12GrowCapacityER5nsStrj+112>: add %eax,%eax
> 0x080714f2 <_ZN12nsStrPrivate12GrowCapacityER5nsStrj+114>: cmp %esi,%eax
> 0x080714f4 <_ZN12nsStrPrivate12GrowCapacityER5nsStrj+116>: jb 0x80714f0
I'm concerned than compilers *can* generate code like this, as it only aids in proving Murphy's
Law: Whatever can go wrong, will. It's a general design flaw that %eax and %esi are not checked
for a working combination b4 this code. It's only fate that allows an interrupt to change the
value mid loop, causing the law to enter effect.
The real solution is to validate the loop's variables mid loop. That way no amount of trickery,
including mathematical errors, will cause an endless loop.
Who besides me is looking for and trying to correct these anomalies?
__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus
Reply to: