|
From: Paulo J. M. <po...@so...> - 2006-11-17 15:50:34
|
Hi all, When running my program (compiled with gcc4.1.1, with -g -Wall --std=c++98) under valgrind 3.2.0 with --leak-check=yes I get error as: ==6856== Invalid read of size 4 ==6856== at 0x804E92F: smtparse() (smtlib-parser.cc:2193) ==6856== Address 0xBEBE8AC0 is on thread 1's stack However, I'm not using threads in my program. Can I get more explanatory error messages? Regards, -- Paulo Jorge Matos - pocm at soton.ac.uk http://www.personal.soton.ac.uk/pocm PhD Student @ ECS University of Southampton, UK |
|
From: Tom H. <to...@co...> - 2006-11-17 19:36:31
|
In message <11b...@ma...>
"Paulo J. Matos" <po...@so...> wrote:
> When running my program (compiled with gcc4.1.1, with -g -Wall
> --std=c++98) under valgrind 3.2.0 with --leak-check=yes I get error
> as:
> ==6856== Invalid read of size 4
> ==6856== at 0x804E92F: smtparse() (smtlib-parser.cc:2193)
> ==6856== Address 0xBEBE8AC0 is on thread 1's stack
>
> However, I'm not using threads in my program.
Thread 1 is the initial thread - the one created at program startup
for any program.
> Can I get more explanatory error messages?
What more would you like it to tell you?
Tom
--
Tom Hughes (to...@co...)
http://www.compton.nu/
|
|
From: Paulo J. M. <po...@so...> - 2006-11-18 16:25:24
|
Thank you for both your replies. In fact, the problem is that it seems really confusing the amount of errors produces by valgrind. In the beginning: ==23293== Warning: client switching stacks? SP change: 0xBEABDB4C --> 0xBE396A00 ==23293== to suppress, use: --max-stackframe=7500108 or greater Is this important? First error is: ==23293== Invalid write of size 4 ==23293== at 0x804D0F6: smtparse() (smtlib-parser.cc:1355) ==23293== Address 0xBEABDB18 is on thread 1's stack This is : int yytoken = 0; This is bison generated code. How can this line contain an invalid write? Am I doing something wrong? I'm running valgrind: valgrind --leak-check=yes program args... I just go through the first 10 invalid writes reported and they all refer to bison generated code. Any ideas on this? Regards, -- Paulo Jorge Matos - pocm at soton.ac.uk http://www.personal.soton.ac.uk/pocm PhD Student @ ECS University of Southampton, UK |