Robin, On 10/13/2018 11:46 AM, Walter Harms wrote: > > >> Robin Kuzmin <kuzmin.robin@xxxxxxxxx> hat am 10. Oktober 2018 um 23:40 >> geschrieben: >> >> >> http://man7.org/linux/man-pages/man3/alloca.3.html >> >> I compare this man page with the >> https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/alloca?view=vs-2017 >> and I feel that a very important fragment is missing in the man page >> (although strictly speaking the man page is complete), the missing >> fragment is between the double asterisks (**): >> >> The allocated space is automatically freed when the calling function >> exits **(not when the allocation merely passes out of scope)**. > > My Version reads: > "This temporary space is automatically freed when the function that called > alloca() returns to its caller." > I thing this is pretty clear. I'm inclined to agree with Walter actually. And indeed I find the text that you quote is actually a bit confused. The *allocation* doesn't pass out of scope; the *pointer* to the allocation might pass out of scope. >> More info: >> * Know the Danger of alloca() >> https://github.com/kuzminrobin/code_review_notes/blob/master/cpp_design_bookmarks.md#know-the-danger-of-alloca There, you say (I assume s/malloca/alloca/): Thus if the malloca() is called in the loop then each loop iteration allocates more and more space on the stack which can easily cause the stack overflow. Such a behavior is fundamentally differnt from the behavior of its counterparts - the ordinary arrays (whose number of elements is specified with a compile-time constant), and the variable length arrays (whose number of elements is specified with a run-time value). Both of these couterparts reuse the space on the stack during each iteration of the loop. But, this is to me a slightly surprising way of thinking about alloca(). To me, the counterparts for behavior are malloc() etc. And if you used malloc() in the same way, again you would have a memory issue (but on the heap, rather than the stack). So, I'm not saying you're wrong in any of your assertions, it's just that (to me) it's a surprising way of (mis)understanding alloca(). But, perhaps some people do (mis) understand alloca() as you suggest, so I've added the following sentence in NOTES: The space allocated by alloca() is not automatically deallocated if the pointer that refers to it simply goes out of scope. Cheers, Michael -- Michael Kerrisk Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/ Linux/UNIX System Programming Training: http://man7.org/training/
