On Sun, 11 Feb 2007 15:12:40 -0800 Mounir Bakkali wrote: > Nish Aravamudan wrote: > > On 2/11/07, Mounir Bakkali <mounir.bakkali@xxxxxxxxx> wrote: > >> Well the kernel manages the memory and malloc makes a kernel call > >> to give you or not the amount of memory requested... so it is a kernel > >> issue and therefore my question fits totally in here. > > > > 1) Don't top-post. > > > > 2) The kernel also manages your filesystem and your hard drives. That > > doesn't make asking about why your hard drive is out of space a kernel > > issue. > > > >> Moreover, I found this on wikipedia: > >> > >> > >> "Allocation failure > >> > >> |malloc| is not guaranteed to succeed — if there is no memory available, > >> or if the program has exceeded the amount of memory it is allowed to > >> reference, |malloc| will return a |NULL| pointer. " > > > > Ahem, someone else already pointed this out to you on the list. And > > it's in the man-page which is where you should look for programming > > information. > > > >> And I am not in these 2 cases so the kernel should be able to allocate > >> 256 bytes while I have nearly > >> 800 Mega bytes of free memory. > >> The thing I cannot figure out is why it is working with 512 (which is > >> bigger than 256....obviously > >> it can allocate more memory but not less???) and not with 256. > >> > >> There is something though I forgot to mention. The program is allocating > >> 256 bytes for about > >> 2500 times but the segmentation fault appears about the 2000th > >> allocation and I know that later on > >> there will be a segmentation fault because later on I need 376 bytes > >> (after the 2050th allocation) > >> and I know that at that time there will be a segmentation fault. > >> > >> So my question is: why is this segmentation fault appears after the > >> 200th and not after the 2000th allocation? > > > > You just said it happens after the 2000th time. Show us the code and > > the exact problem, if you want any help. But again, this is not a > > problem for discussion on this list. > > > >> Does it have to do with some kind of locking features and memory of the > >> kernel? Can it be a scheduling issue? > > > > This is a list for asking about kernel *development*. Not userspace > > programming issues. > > > > This is not to say you may not have run into a kernel bug. But it > > seems highly unlikely, as malloc() is so widely used and no one else > > is complaining. > > > > Thanks, > > Nish > > > > 1) You answers do not help at all Nish. Talking around the topic do not > help. > > 2) Nish, I am not complaining about anything. I am just trying to understand > and eventually to help the open-source industry. > > 3) memory management is a hot topic regarding the linux kernel. Sure, but so far your example is a userspace problem, most likely glibc. > 4) For you Nish exclusively: Being nice does not require much effort > so please, do not answer reply regarding this topic since it looks like > you prefer to talk around the topic instead of helping. > > 5) Kernelnewbies: I apologize for this email. I don't like to write such > email > but I had to because of that Nish guy. > Please if anyone has any idea, especially if it is really a kernel > issue, feel > free to post it in here. As already suggested, try to use gdb to find the fault. And at last post a full example program so that other people can try it out. --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
