The Glade 4.0

"Turn the lights down, the party just got wilder."
It is currently Sun Nov 24, 2024 1:19 pm

View unanswered posts | View active topics


Board index » The Glade » Tech

All times are UTC - 6 hours [ DST ]



Post new topic Reply to topic  Page 1 of 1
  [ 9 posts ] 
  Print view Previous topic | Next topic 
Author Message
Lex Luthor
 Post subject: C Programming Question
PostPosted: Wed Feb 09, 2011 11:29 pm 
Offline

Joined: Thu Sep 03, 2009 10:03 am
Posts: 4922
I figured out that the following changes were necessary to stop the process from crashing. Does anyone know why? It's especially odd because in one build it works fine, but for another it leads to a runtime crash with a seg fault. The part that caused a segfault is &pool_info[loop_cnt] from right below. Maybe I'm just missing something obvious.

Code:
          rc = show_pool_brief_get_bag(svr_datalist,
                                               &pool_info[loop_cnt]);


Here's the relevant part of the diff:

Spoiler:
Code:
  uint32_t num_addrs_free = 0;
      uint32_t num_addrs_excl = 0;
      uint32_t num_addrs_total = 0;
!     ip_show_pool_brief pool_info[MAX_SHOW_POOLS];
 
      memset(&util, 0, sizeof(util));
 
      rc = show_sysdb_bind_server(&server_edm);
      if (CERR_IS_NOTOK(rc)) {
--- 271,280 ----
      uint32_t num_addrs_free = 0;
      uint32_t num_addrs_excl = 0;
      uint32_t num_addrs_total = 0;
!     ip_show_pool_brief *pool_info[MAX_SHOW_POOLS];
 
      memset(&util, 0, sizeof(util));
+     memset(&pool_info, 0, sizeof(pool_info));
 
      rc = show_sysdb_bind_server(&server_edm);
      if (CERR_IS_NOTOK(rc)) {
***************
*** 289,307 ****
      }
 
      for (loop_cnt = 0; loop_cnt < svr_bag_cnt; loop_cnt++) {
          rc = show_pool_brief_get_bag(svr_datalist,
!                                           &pool_info[loop_cnt]);
          if (CERR_IS_NOTOK(rc))
          {
              svr_bag_cnt = loop_cnt;
              break;
          }
!         num_addrs_alloc += pool_info[loop_cnt].num_addrs_alloc;
!         num_addrs_free += pool_info[loop_cnt].num_addrs_free;
!         num_addrs_excl += pool_info[loop_cnt].num_addrs_excl;
!         num_addrs_total += pool_info[loop_cnt].num_addrs_alloc +
!             pool_info[loop_cnt].num_addrs_excl +
!             pool_info[loop_cnt].num_addrs_free;
      }
      util.total_util_current = (num_addrs_alloc * 100) / num_addrs_total;
         
--- 290,314 ----
      }
 
      for (loop_cnt = 0; loop_cnt < svr_bag_cnt; loop_cnt++) {
+         pool_info[loop_cnt] = calloc(1, sizeof(ip_show_pool_brief));
+         if (!pool_info[loop_cnt]) {
+             printf("Could not allocate memory");
+             svr_bag_cnt = loop_cnt;
+             goto end;
+         }
          rc =  show_pool_brief_get_bag(svr_datalist,
!                                          pool_info[loop_cnt]);
          if (CERR_IS_NOTOK(rc))
          {
              svr_bag_cnt = loop_cnt;
              break;
          }
!         num_addrs_alloc += pool_info[loop_cnt]->num_addrs_alloc;
!         num_addrs_free += pool_info[loop_cnt]->num_addrs_free;
!         num_addrs_excl += pool_info[loop_cnt]->num_addrs_excl;
!         num_addrs_total += pool_info[loop_cnt]->num_addrs_alloc +
!             pool_info[loop_cnt]->num_addrs_excl +
!             pool_info[loop_cnt]->num_addrs_free;
      }
      util.total_util_current = (num_addrs_alloc * 100) / num_addrs_total;
Top
 Profile  
Reply with quote  
Stathol
 Post subject:
PostPosted: Thu Feb 10, 2011 10:22 am 
Offline
Lean, Mean, Googling Machine
User avatar

Joined: Thu Sep 03, 2009 9:35 am
Posts: 2903
Location: Maze of twisty little passages, all alike
It's hard to say, being unfamiliar with the code. What is the declaration for show_pool_brief_get_bag()?

In the first version of the code, you're passing it the address of pool_info[loop_cnt] (which is the same as pool_info + (loop_cnt * sizeof(pool_info[0])), if you think about it.) in the second, you're passing it the value pointed to by pool_info[loop_cnt]. What is the function expecting?

_________________
Sail forth! steer for the deep waters only!
Reckless, O soul, exploring, I with thee, and thou with me;
For we are bound where mariner has not yet dared to go,
And we will risk the ship, ourselves and all.
Top
 Profile  
Reply with quote  
Lex Luthor
 Post subject: Re:
PostPosted: Sun Feb 13, 2011 5:33 pm 
Offline

Joined: Thu Sep 03, 2009 10:03 am
Posts: 4922
Stathol wrote:
It's hard to say, being unfamiliar with the code. What is the declaration for show_pool_brief_get_bag()?

In the first version of the code, you're passing it the address of pool_info[loop_cnt] (which is the same as pool_info + (loop_cnt * sizeof(pool_info[0])), if you think about it.) in the second, you're passing it the value pointed to by pool_info[loop_cnt]. What is the function expecting?


Sorry for the late reply. show_pool_brief_get_bag() takes in the same types of argument that are passed to it. It compiles fine. It expects this type: ip_show_pool_brief *.I tried wiping out the function so that it simply returns upon entering, but it would still crash. It turned out the problem is not in show_pool_brief_get_bag().

I'm just going to consider the cause of this problem unsolveable :P. My fix worked though.
Top
 Profile  
Reply with quote  
Rorinthas
 Post subject:
PostPosted: Tue Feb 15, 2011 9:51 am 
Offline
pbp Hack
User avatar

Joined: Wed Sep 02, 2009 8:45 pm
Posts: 7585
Welcome to the world of troubleshooting! Sometimes fixes just work and there is no good, logical, apparent, explanation for it.

_________________
I prefer to think of them as "Fighting evil in another dimension"
Top
 Profile  
Reply with quote  
Mookhow
PostPosted: Tue Feb 15, 2011 5:47 pm 
Offline
God of the IRC
User avatar

Joined: Wed Sep 02, 2009 7:35 pm
Posts: 3041
Location: The United States of DESU
Once I was helping someone troubleshoot his C program. The program kept crashing when trying to manipulate a certain variable (divide by zero?). I added a print statement higher up in the code to verify the value of the variable before trying to manipulate it. The problem went away. I took out the printf(), and the problem came back. We ended up leaving an empty printf() statement in the code.

_________________
Image
Top
 Profile  
Reply with quote  
Lex Luthor
PostPosted: Tue Feb 15, 2011 7:59 pm 
Offline

Joined: Thu Sep 03, 2009 10:03 am
Posts: 4922
Mookhow wrote:
Once I was helping someone troubleshoot his C program. The program kept crashing when trying to manipulate a certain variable (divide by zero?). I added a print statement higher up in the code to verify the value of the variable before trying to manipulate it. The problem went away. I took out the printf(), and the problem came back. We ended up leaving an empty printf() statement in the code.


That sounds like highly unstable code. :P It could break if you try to port it to another platform.
Top
 Profile  
Reply with quote  
Stathol
 Post subject:
PostPosted: Tue Feb 15, 2011 11:55 pm 
Offline
Lean, Mean, Googling Machine
User avatar

Joined: Thu Sep 03, 2009 9:35 am
Posts: 2903
Location: Maze of twisty little passages, all alike
http://en.wikipedia.org/wiki/Unusual_software_bug#Heisenbug

_________________
Sail forth! steer for the deep waters only!
Reckless, O soul, exploring, I with thee, and thou with me;
For we are bound where mariner has not yet dared to go,
And we will risk the ship, ourselves and all.
Top
 Profile  
Reply with quote  
Timmit
 Post subject:
PostPosted: Wed Feb 16, 2011 10:33 am 
Offline

Joined: Sun Sep 06, 2009 12:09 pm
Posts: 733
I used to maintain some fortran database code on an old Sperry Unisys mainframe that had an odd comment in the middle of one of it's subroutines that said only "Do not remove this comment or the program will break".

Of course, that's impossible, right? So we deleted the comment and suddenly it wouldn't compile...

We put the comment back and it compiled again. Computers are magical...
Top
 Profile  
Reply with quote  
Lex Luthor
 Post subject:
PostPosted: Wed Feb 16, 2011 10:38 am 
Offline

Joined: Thu Sep 03, 2009 10:03 am
Posts: 4922
I've been learning that while compilers and platforms seem to be dependable 99% of the time, they aren't perfect.
Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  Page 1 of 1
  [ 9 posts ] 

Board index » The Glade » Tech

All times are UTC - 6 hours [ DST ]

Who is online

Users browsing this forum: No registered users and 136 guests

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group