Bug#720352: fwrite: crash upon failed write(2)

[Stephane Chazelas <stephane.chazelas@gmail.com>]

I'd say there's at least one bug and it doesn't seem to be fixed
upstreams
(http://sourceware.org/git/?p=glibc.git;a=blob;f=libio/fileops.c;h=e92f85b243496c07d3677b97c785da7f42fb6c38;hb=HEAD#l531)
so you may want to forward to them:

 530   count = _IO_SYSWRITE (fp, data, to_do);
 531   if (fp->_cur_column && count)

That should probably be "&& count > 0" above. And that would explain the 8192 that becomes a 8193.

It may not be the only problem because I can see a flag is set
upon the error of write(2) but it does not seem to be checked on
the write path (from a very quick cscope check).

-- 
Stephane