On Wed, 2014-07-30 at 11:56 +0200, Carl-Johan Kjellander wrote: > On Wed, Jul 30, 2014 at 11:09 AM, Gisle Vanem <gva...@yahoo.no> wrote:
> Sure, I can try. Latest git master: > http://pastebin.com/4UEj0zYm > So bug is still there. However... Can you give some information on the type of hardware and operating system you're using? > While searching for instructions on how to compile 'make' (the INSTALL > FILE doesn't really say that you need to run 'make update') The INSTALL file is only for building GNU make from a distributed source tarball, say from ftp.gnu.org. If you want to build from Git you must read the README.git file and follow those directions. > Any ideas about the original bug though? That message means that while make was printing information to stdout (note that the only things make itself prints to stdout during a normal runtime are the "entering/exiting" lines, plus the command line it will invoke for each rule where it isn't suppressed), the fprintf() command got an error and wasn't able to write its output properly. This error comes from the C runtime library and denotes some kind of error writing content to stdout. I don't see, offhand, how there could be any bug in GNU make that would cause this problem. Is there anything unique or unusual about the way you're invoking make, with respect to its stdout? Are you writing the output to a file using redirection? Reading the output using a pipe (piping make's output to another program)? Something else? _______________________________________________ Bug-make mailing list Bug-make@gnu.org https://lists.gnu.org/mailman/listinfo/bug-make
