Changelog
flex (2.5.35-3ubuntu1) jaunty; urgency=low
* Merge from debian unstable, remaining changes:
- Process (but ignore) the fwrite used in the "ECHO" macro to handle
being compiled with -D_FORTIFY_SOURCE=2 (debian bug 505233).
flex (2.5.35-3) unstable; urgency=low
* Bug fix: "debian dir contains *.so", thanks to Raphael Hertzog
This is a simple rname of a file used purely in packaging, and has low
risk, but would allow compiling the flex source package with squeeze's
packaging tools. (Closes: #485095).
* [88bb2fc] Fix generation of broken output on ia64 when input is piped in
On ia64, flex appears to generate a completely different (and broken)
output depending on whether the input file is specified using shell
redirection or via a pipe.
This was traced back commit 3971e1a917548977cff71418a7c3575ffbc9571f,
by Alex Nixon <email address hidden>: VFS: increase pseudo-filesystem
block size to PAGE_SIZE. PAGE_SIZE on ia64 kernels is larger than most
architectures - 16K is what most distributions (including Debian)
use. An effect of the above changeset, as verified by strace, is that
the read() and write() system calls issued by flex on a pipe will use
this bigger size. fread() appears to be properly hiding the actual
read size and is returning 8192 byte buffers to flex upon request. I
verified by this by instrumenting YY_INPUT to compare the fread data
with the original input file.
Its probably useful to note that the corruption happening here is that
the last 382 lines of the input file is being prefixed to the output
file. The only differences in the remainder of the file are the #line
numbers. I've also found that bumping YY_READ_BUF_SIZE up to 16384
seems to workaround the problem - but that may just be putting it back
into hiding.
The tradeoff here is between correct operation, by making flex's
buffer at least as big as the system buffer -- and not bigger, so that
we do not waste memory. At this point, fixing the isue seems more
important than worrying that flex consumes and additional 8KB of
memory.
This fix unblocks RC bugs on other packages.
Bug fix: "Generates broken output on ia64 when input is piped vs. shell
redirection", thanks to Chris Lamb. Fix based on comments by Dann
Frazier and Michael Casadevall. (Closes: #501960)
* [56ec97d] Fix doxygen comments to reflect actual argument name
Fix the comments to reflect the actual name of the function
arguments. This is a pure documentation fix.
Bug fix: "argument 'bytes' of command @param is not found in
the argument list of tp_scan_bytes(yyconst char *yybytes, int
_yybytes_len)", thanks to Ludovic Rousseau (Closes: #488272).
-- Kees Cook <email address hidden> Mon, 10 Nov 2008 15:25:58 -0800