Joerg Schilling wrote: > Guy Shaw <Guy.Shaw at Sun.COM> wrote: > >> I mentioned that I recently added the system call, kprintf(). > > > Why? > What is the reson for such a syscall?
During early bringup, especially before rtld+libc gets debugged and fully initialized, it is about the _only_ practical way to get any information out to the console. I introduced the topic in an early message entirled, "Polaris status -- rtld", in the section called "Can't see". But, I should not expect everyone to wade through all my earlier messages. I don't have it worked out how much context I should supply in each message to make them more self-contained. >> Kernel >> ====== >> [ ] System call numbers are in >> usr/src/uts/common/sys/syscall.h >> >> [ ] For kprintf(), the syscall number 56 is available, apparently. > > > Why do you believe that 56 is available? Because I looked at the source code, and it was the only one that was not used. William Kucharski answered this, as well. Thanks. > > In former times, /usr/include/sys/syscall.h did list reasons > for unavailable numbers. I have no idea where to get this from now. > It is most unlikely that syscall # 56 is "available". > > J?rg Don't worry. kprintf() will _never_ make it into Solaris product. -- Guy Shaw
