# Rb\_raise problem in C extension **URL:** \ **Category:** Ruby\ **Created:** [April 22, 2009, 7:07pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206 "2009-04-22T19:07:04Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 22, 2009, 7:07pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/1 "2009-04-22T19:07:04Z")
I am using Ruby 1.8.6-26 from the One Click Installer on Windows. I have a C extension that tries to calloc() memory. If the calloc() fails I call rb\_raise(rb\_eNoMemError,“Cannot allocate data”). My program is getting stuck in this code. Debugging (unfortunately via print statements) I can see that rb\_raise() is going to be called. After that the exception is never caught by the outermost rescue loop. The program just stops doing anything (0% cpu) except it keeps updating a timer in another thread. Are there things I need to know about rb\_raise() and how to use it? ---
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 22, 2009, 8:46pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/2 "2009-04-22T20:46:29Z")
On Apr 22, 2009, at 10:05, removed\_email\_address@domain.invalid wrote: > I am using Ruby 1.8.6-26 from the One Click Installer on Windows. I > have a C extension that tries to calloc() memory. If the calloc() > fails I call rb\_raise(rb\_eNoMemError,“Cannot allocate data”). My > program is getting stuck in this code. Debugging (unfortunately via > print statements) I can see that rb\_raise() is going to be called. > After that the exception is never caught by the outermost rescue loop. > The program just stops doing anything (0% cpu) except it keeps > updating a timer in another thread. Are there things I need to know > about rb\_raise() and how to use it? If ruby is out of memory how could it allocate more memory to raise an exception? Ruby itself allocates a NoMemError at startup to ensure it can raise one when it runs out of memory. You’ll probably need to do the same. See gc.c rb\_memerror(). ---
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 22, 2009, 9:56pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/3 "2009-04-22T21:56:05Z")
On Apr 22, 1:44 pm, Eric H. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > about rb\_raise() and how to use it? > > If ruby is out of memory how could it allocate more memory to raise an > exception? > > Ruby itself allocates a NoMemError at startup to ensure it can raise > one when it runs out of memory. You’ll probably need to do the same. > See gc.c rb\_memerror(). There is plenty of memory available (4-6 gig free). But I’m asking calloc() for a 1 gb chunk and it can’t find one. Is there a different exception I should throw in this case? Will the rb\_raise() in my nested C code percolate out to my handler in my nested ruby code? (As a test for now I’m just changing it to an eException but would appreciate any feedback you have) ---
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 23, 2009, 7:41pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/4 "2009-04-23T19:41:01Z")
On Wed, 22 Apr 2009 13:44:52 -0500, Eric H. wrote: > If ruby is out of memory how could it allocate more memory to raise an > exception? > > Ruby itself allocates a NoMemError at startup to ensure it can raise one > when it runs out of memory. You’ll probably need to do the same. See > gc.c rb\_memerror(). Is rb\_memerror() exposed for him to call? He could just call that, and it would spare him all issues with preallocation. –Ken ---
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 23, 2009, 12:04am UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/5 "2009-04-23T00:04:33Z")
On Apr 22, 2009, at 12:55, removed\_email\_address@domain.invalid wrote: > > > updating a timer in another thread. Are there things I need to know > > > There is plenty of memory available (4-6 gig free). But I’m asking > > > calloc() for a 1 gb chunk and it can’t find one. Is there a different > > > exception I should throw in this case? Will the rb\_raise() in my > > > nested C code percolate out to my handler in my nested ruby code? (As > > > a test for now I’m just changing it to an eException but would > > > appreciate any feedback you have) In that case, rb\_raise should do what you want, however you may need to explicitly rescue it: $ ruby begin begin raise NoMemoryError rescue puts “caught with plain rescue” end rescue Exception # or NoMemoryError puts “caught with rescue Exception” end ^D caught with rescue Exception For this reason, you should use RuntimeError or StandardError (especially for custom error classes) instead of Exception, since Exception isn’t caught by a plain rescue. ---
**Author:** ![unknown](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/u/4bbf92/32.png) [@unknown](https://www.ruby-forum.com/u/unknown)\ **Post date:** [April 23, 2009, 7:50pm UTC](https://www.ruby-forum.com/t/rb-raise-problem-in-c-extension/166206/6 "2009-04-23T19:50:23Z")
On Apr 23, 2009, at 10:40, Ken B. wrote: > > > the exception is never caught by the outermost rescue loop. The > > > Ruby itself allocates a NoMemError at startup to ensure it can > > > raise one > > > when it runs out of memory. You’ll probably need to do the same. See > > > gc.c rb\_memerror(). > > Is rb\_memerror() exposed for him to call? He could just call that, > and it > would spare him all issues with preallocation. yeah, looks like it’s in intern.h.