I've seen once on a site, a call graph digging into the most low level libraries of a web request initiated by a PHP appplication with call timings and summary. Seems to me that this is a great way to spot the reason of bottlenecks that aren't obvious while profiling PHP-only code.
我曾经在网站上看过一次调用图,这个调用图挖掘了一个由PHP应用程序启动的Web请求的最低级库,其中包含调用时间和摘要。在我看来,这是一个很好的方法来发现在分析PHP专用代码时不明显的瓶颈的原因。
Something like strace does but with far more detail.
像strace这样的东西确实有更多细节。
2 个解决方案
#1
Rasmus mentioned valgrind a couple times in his presentations. If I remember correctly it measures IO, memory, etc. Might want to take a look.
拉斯穆斯在他的演讲中曾多次提到过valgrind。如果我没记错的话,它会测量IO,内存等。可能要看看。
#2
Tools like VTune helps here. You can point this tool at a process or a system to profile and it will help find performance hotspots. VTune also has a call graph mode.
像VTune这样的工具在这里有帮助您可以将此工具指向流程或系统以进行配置,这将有助于查找性能热点。 VTune还有一个调用图模式。
#1
Rasmus mentioned valgrind a couple times in his presentations. If I remember correctly it measures IO, memory, etc. Might want to take a look.
拉斯穆斯在他的演讲中曾多次提到过valgrind。如果我没记错的话,它会测量IO,内存等。可能要看看。
#2
Tools like VTune helps here. You can point this tool at a process or a system to profile and it will help find performance hotspots. VTune also has a call graph mode.
像VTune这样的工具在这里有帮助您可以将此工具指向流程或系统以进行配置,这将有助于查找性能热点。 VTune还有一个调用图模式。