__
CIL-users mailing list
CIL-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cil-users
From f7ca8992341395c83e2a6c0a0986eeedc4b8fbf2 Mon Sep 17 00:00:00 2001
From: Oliver Schwahn
Date: Mon, 18 Aug 2014 15:14:10 +0200
Subject: [PATCH] Added support f
Hi,
I can confirm that the '--keepunused' option is needed to trigger this
behavior. Without it, CIL produces the expected output, i.e., no
parameter names are replaced in the function definition.
I submitted a bug report on SF (#148).
Cheers,
Oliver
On 2014-08-15 21:52, Gabriel Kerneis wrot
---
** [bugs:#148] Wrong parameter name substitution**
**Status:** open
**Group:** Bug
**Created:** Mon Aug 18, 2014 10:13 AM UTC by Oliver Schwahn
**Last Updated:** Mon Aug 18, 2014 10:13 AM UTC
**Owner:** nobody
If a function definition is followed by a function prototype that uses
Hi,
I just noticed that CIL (the git develop branch version) messes up
functions if the function definition is followed by a function prototype
that uses different names for parameters. CIL replaces the actual
function parameter names with the parameter names of the prototype.
Please see the f
Hi,
> So, what I'm currently using CIL for in my tool is to preprocess the
> source file under analysis to make it easier to analyze -- or at
> least, that's the hope.
Ok, so you just want the merged C source as output? No compilation or
linking?
> I'm interfacing with Cilly through Python. T
Hi,
> I was wondering if there was a command-line option that can be passed
> to the merger that mimics setting this environment variable.
>
No, there is no such command line option that i am aware of. But...
The compilation and linking control is implemented in cilly (Perl
script) which mimics
Hi,
I'm not sure if I fully understand your question. But some time ago,
there was another question regarding the merger where I posted an
archive containing a small example C project along with some scripts
which invoke the merger in different ways. Maybe this helps you with
your problem.
He
Hi,
I also think that this is invalid C code. Initializers for static local
variables have to be constant values. Variables with const modifier are
not considered a constant value in this context (at least not by gcc,
not sure what the standard says).
So, the following code does also NOT work..
Hi,
I attached an archive which contains your test project along with some
script files which show how you can use cilly for merging.
The merge scripts are pretty simple. They just invoke cilly in some
different ways to obtain the merged source and the compiled program.
This should give you an
Ah, ok. Yours is better and works just fine.
Cheers
Oliver
19.07.2012 10:40, Gabriel Kerneis:
> Yes, it has been reported already. I did a patch yesterday but forgot to push
> it, sorry.
>
> http://github.com/kerneis/cil/commit/b8089965ff735aed1863cd19409e14b8d39cbe01
>
> It is slightly differen
er
14.07.2012 08:41, Gabriel Kerneis:
On Fri, Jul 13, 2012 at 01:49:39PM +0200, Oliver Schwahn wrote:
I just realized that the CIL build chain does not support the native
versions of the OCaml tools. The native tool versions often
outperform the byte code versions, e.g. ocamlc.opt (the native
version)
Hi,
I just realized that the CIL build chain does not support the native
versions of the OCaml tools. The native tool versions often outperform
the byte code versions, e.g. ocamlc.opt (the native version) can compile
CIL much faster than ocamlc (the byte code version) can.
I consider support
Hi,
I tested the patch and it seems to work fine! I encountered no problems
at all.
Cheers,
Oliver
--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape
Hi Gabriel,
> I'm afraid commenting out wouldn't work (well not sure, but it could be
> stripped
> too early in preprocessing passes), but I just realised we could use a string
> instead of an integer?
I think using a string here seems to be the best option.
> Please, could you try the attached
Hi!
>> A quick workaround is to set the system clock to a time before 2004.
> Or to use a 64-bit architecture (yielding max_int = 2^63-1).
You are right, using a 64-bit architecture is, of course, also a
workaround but I have to stick to my 32-bit system :-(
>> I think this issue can be fixed ei
Hi all,
I just tried to use the merger (together with cilly) but i noticed that
it fails with error message:
Unimplemented: something.c:1: Int constant too large: 1341229737
It turns out that the problem is the "#pragma merger (...)"which is
written by cilly as first line of the merged output
Hi!
Thank you very much! The patch works perfectly!
Best regards,
Oliver Schwahn
--
Virtualization & Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing
Hi,
I have a question concerning function parameters.
Why does CIL remove the const qualifier on all function parameters with
non-pointer type?
Here is an example:
// original function
void all_const(const int a0, const long a1, const char a2)
{ ... }
// CIL output
void all_const(int a0 , long a
18 matches
Mail list logo
