Are there any speed/memory implications for using "always dbstop on error"?

Question

Richard Crozier 2011-5-18

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/7651-are-there-any-speed-memory-implications-for-using-always-dbstop-on-error

I often find it handy to use the "Always dbstop on error" option for the debugger, but am wondering if there is any drawback in terms of speed and memory to always using this option. I have no particular reason to think that there is, but am just wondering if there is a reason, other than personal preference not to have this setting as the default in my startup.m file.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Jan 2011-5-18

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/7651-are-there-any-speed-memory-implications-for-using-always-dbstop-on-error#answer_10574

The JIT acceleration does not gain full speed in debug mode. With the JIT compilation the order of executions can be influenced, such that line-by-line debugging would lead to surprising results.

Therefore a breakpoint in the code slows down the execution measureable, 10% in some tests. The effects of "dbstop if error" are smaller, about 1 to 2% in my tests. But this is very near to the measurement accuracy. So I assume the benefits for debugging will top te drawback for the speed.