Re: Strange intermittent EIO error when writing to stdout sincev3.8.0

[Markus Trippelsdorf]

On 2013.06.06 at 18:31 +0100, David Howells wrote:
> Markus Trippelsdorf <markus@xxxxxxxxxxxxxxx> wrote:
> 
> > OSError: [Errno 5] Input/output error
> > 
> > Basically 'emerge' just writes the build output to stdout in a loop:
> 
> I've just upgraded to kernel-3.9.4-200.fc18.x86_64 and I'm now regularly
> seeing something very similar from the tee run by "fedpkg local" to log its
> output when building a cross-compiler gcc RPM:
> 
> 	tee: standard output: Input/output error
> 	tee: write error
> 
> Slapping strace on tee after the "Input/output error" line has occurred shows
> that tee is still running and still pushing lines from its stdin to the output
> named on its command line, but has ceased writing to its stdout.

tee uses the same loop as portage, but keeps going in case of an EIO:

  /* In the array of NFILES + 1 descriptors, make
     the first one correspond to standard output.   */
  descriptors[0] = stdout;
  files[0] = _("standard output");
  setvbuf (stdout, NULL, _IONBF, 0);

  for (i = 1; i <= nfiles; i++)
    {
      descriptors[i] = (STREQ (files[i], "-")
                        ? stdout
                        : fopen (files[i], mode_string));
      if (descriptors[i] == NULL)
        {
          error (0, errno, "%s", files[i]);
          ok = false;
        }
      else
        setvbuf (descriptors[i], NULL, _IONBF, 0);
    }

  while (1)
    {
      bytes_read = read (0, buffer, sizeof buffer);
      if (bytes_read < 0 && errno == EINTR)
        continue;
      if (bytes_read <= 0)
        break;

      /* Write to all NFILES + 1 descriptors.
         Standard output is the first one.  */
      for (i = 0; i <= nfiles; i++)
        if (descriptors[i]
            && fwrite (buffer, bytes_read, 1, descriptors[i]) != 1)
          {
            error (0, errno, "%s", files[i]);
            descriptors[i] = NULL;
            ok = false;
          }
    }

-- 
Markus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/