On Fri, Nov 16, 2012 at 04:40:50PM +0100, Virgile Prevosto wrote:
> Another solution is to enclose the statement in a block and use a
> block attribute (field battrs)
Excellent idea, in particular because block attributes do not need to be unique
in a function scope (contrary to labels).
--
Gabr
Hello,
2012/11/16 Gabriel Kerneis :
> On Thu, Nov 15, 2012 at 06:01:50PM +0100, Pietro Braione wrote:
>> Our current idea is to inject some kind of assignment statement to a fubar
>> variable as, e.g., _next_statement_is_relevant = 1, right before every
>> "relevant" statement. Can anyone suggest
Il giorno 16/nov/2012, alle ore 14.29, Gabriel Kerneis ha scritto:
> On Thu, Nov 15, 2012 at 06:01:50PM +0100, Pietro Braione wrote:
>> Our current idea is to inject some kind of assignment statement to a fubar
>> variable as, e.g., _next_statement_is_relevant = 1, right before every
>> "relevant
On Thu, Nov 15, 2012 at 06:01:50PM +0100, Pietro Braione wrote:
> Our current idea is to inject some kind of assignment statement to a fubar
> variable as, e.g., _next_statement_is_relevant = 1, right before every
> "relevant" statement. Can anyone suggest something better?
Some magic label on rel
Hello to everyone. We are using CIL in one of our tools for automated test
generation. We run a number of CIL passes which normalize and finally
instrument the C code under analysis. The problem is, there are some statement
in the input source that are "relevant", we want to transmit this inform
