On Sun, 1 Aug 2010, Eduardo Horvath wrote:
Why is there a need to hold a mutex across module loading instead of, say, having a reference count tht's protected by a mutex? I thought holding a mutex across a potentially blocking operation such as a module load is considered bad practice.
Please note that I am not introducing the holding of the lock over the potentially blocking operation. This is already done.
Several others have already indicated that the atomicity of the module load (including the loading of any required modules) is a desirable property. I'm simply looking for a mechanism that allows us to keep this property while allowing modules to load other modules.
------------------------------------------------------------------------- | Paul Goyette | PGP Key fingerprint: | E-mail addresses: | | Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com | | Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net | | Kernel Developer | | pgoyette at netbsd.org | -------------------------------------------------------------------------