| |||||||||||||
Signed overflow optimization hazards in the kernelSigned overflow optimization hazards in the kernelPosted Aug 16, 2012 23:10 UTC (Thu) by klossner (subscriber, #30046)In reply to: Signed overflow optimization hazards in the kernel by cmccabe Parent article: Signed overflow optimization hazards in the kernel
Sure, a good programmer should have done that. But the C compiler is often presented with automatically-generated code that has never been seen by human eyes, for example after several levels of macro expansion have done their work. Folding out the resulting dead code is often a substantial win.
(Log in to post comments)
|
|||||||||||||