From: Nobuo Iwata > Sent: 30 September 2016 06:44 > This patch removes checking of socket descriptor value in daemons. > > It was checked to be less than FD_SETSIZE(1024 usually) but it's not > correct. > > To be exact, the maximum value of descriptor comes from > rlimit(RLIMIT_NOFILE). > > Following kernel code determines the value : > get_unused_fd_flags() : fs/files.c > __alloc_fd() : fs/files.c > expand_files() : fs/files.c > > The defalut (soft limit) is defines as INR_OPEN_CUR(1024) in > include/linux/fs.h which is referenced form INIT_RLIMS in > include/asm-generic/resource.h. The value may be modified with ulimt, > sysctl, security configuration and etc. > > With the kernel code above, when socket() system call returns positive > value, the value must be within rlimit(RLIMIT_NOFILE). No extra > checking is needed when socket() returns positive. I'm guessing that the problem is that the code wants to use select() on the socket. The value of RLIMIT_NOFILE isn't really relevant. If fd >= FD_SETSIZE then you can't use select. In reality any program that has anywhere near that number of open fd would be better off using something other than select(). David -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
