Hi Manish and Greg and thank you; This problem had been bothering me for a long time. Your answers were fairly simple and straight forward, but it has been like pulling hen's teeth to get someone to understand the question. If you are ever writing a kernel manual, take a second to outline why the parent-child modal is used, rather than just making a bald statement that that is how Linux does it -- not "Just 'cause I say so" !! On Sat, 2008-05-24 at 15:39 -0400, Greg Freemyer wrote: > On Sat, May 24, 2008 at 2:58 PM, William Case <billlinux@xxxxxxxxxx> wrote: > > Thank you Manish; > > > > I think I am starting to approach an answer I hadn't thought of. Let me > > try it out below. Please correct me if I am wrong. > > > > > > On Sat, 2008-05-24 at 23:40 +0530, Manish Katiyar wrote: > > [snip] > > > >> I don't know about other OS's, but swapper is *handcrafted* using the > >> macro INIT_TASK() macro. > > > > If anyone can confirm that there are other OS's that initiate prcesses > > differently from the Linux parent-child process. No details are really > > needed, I would just like to satisfy myself that either all OS's > > basically use the parent-child modal or that there are several different > > ways that it can be done. For example, would the Windows registry > > system be an alternative?? > > > >> Suppose if we had to start all the processes > >> from raw probably setting the initial values for the task just as > >> swapper will take some time, and you might want to use some kind of > >> template from you can derive the default values of the raw process > >> quickly.....isn't it ??? ....... And isn't it more beneficial if you > >> get the values from the process which is *most* closest in nature to > >> yours so that you have to change only the few key values like tgid, > >> ppid etc. > >> > > To test my understanding by restating: > > > > At boot time, when starting PID 0 (the swapping process) some?, > > several?, all? of the environmental CONSTANTS for my particular system > > are loaded and initialized. By passing those constant values on through > > a parent-child modal to each new process those new processes don't have > > to re-initialize those constants. > > [snip] > Bill,> Is this true? And, might there be other reasons to use a parent-child > > modal? > > > >> Ohh and isn't it the same parent child relationship that we were > >> discussing, or in other words it is a raw process getting default > >> values from a closest template ? > >> > > > > As stated above, this sounds close to a BINGO. > >> > > > > [snip > > > I also think it is useful to realize that UNIX was basically designed > for systems that have a MMU even though low-end systems in the late > 70s / early 80s did not have them. > > I believe there were implementations that ran on 286 based hardware > without MMUs way back then, but they were very kludgy and definately > not the design target for UNIX. > > With an MMU even if a multi-megabyte app is forked, a majority the > memory pages are initially shared between the parent and child. A > copy on write system is used to create individual RAM pages if either > the parent or the child writes to a specific shared RAM page. > > Thus the fork call is very efficient and effectively does the minimum > amount of work necessary to create a new process from a parent > process. Therefore if exec is immediately called by the child, little > wasted effort was caused by it being two discrete steps as opposed to > one larger step. > > And as someone else said with current implementations you have a lot > of control of exactly what is carried over from parent to child via > the clone call and thus can optimize the process. > > Greg -- Regards Bill -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
