On Tue, Feb 14, 2017 at 10:19 AM, Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx> wrote: > On Tue, Feb 14, 2017 at 10:03:08AM -0700, Sam Povilus wrote: >> On Tue, Feb 14, 2017 at 9:36 AM, Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx> wrote: >> > >> > On Tue, Feb 14, 2017 at 08:59:08AM -0700, Sam Povilus wrote: >> > > The number of uartlites should be set by a kernel parameter instead of >> > > using a #define. This allows the user to set the number of uartlites >> > > using only kconfig and not modifying kernel source. >> > > >> > > The uartlite is used by FPGAs that support a basically unlimited number >> > > of uarts so limiting it at 16 dosn't make sense as users might need more >> > > than that. >> > >> > Shouldn't you bound the size here? What happens if you ask for 10000 >> > uarts? >> > >> > thanks, >> > >> > greg k-h >> >> Theoretically there is no limit except FPGA hardware and who knows >> what FPGA designers are thinking. From my understanding fom a kernel >> standpoint it changes the number of "struct uart_port"s declared >> and therefore the amount of memory used by the module, and how >> much time it takes to do lookup. > > There are other internal issues when you try to create zillions of uart > devices, so please, set a sane boundry for this. > > How many uarts did you test this patch with? > > thanks, > > greg k-h I think 256 should be a fairly safe for the next 10 years limit. Does that sound reasonable to you? I tested it with 13 and 16, the hardware I have only has 16 interrupt lines and I didn't see any value in testing something with layered interrupts that dosn't deal with that section of the kernel. -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
