Messages in this thread | | | Date | Sun, 25 Sep 2011 23:02:47 +0800 | Subject | Re: Seems the comment of find_next_system_ram() is not exact | From | Wizard <> |
| |
Borislav
Thanks for your reply and for your suggestion to use "git annotate". I learned a lot from you.
I see the commit by KAMEZAWA. While still not know why he change code this way to cover the overlap case.
I find the function is introduced in 2842f11419704f8707fffc82e10d2263427fc130. While the mm/memory_hotplug.c is chaged during this period. If I want to view the code at that moment, I should use git checkout 2842f11419704f8707fffc82e10d2263427fc130?
2011/9/25, Borislav Petkov <bp@alien8.de>: > On Sat, Sep 24, 2011 at 09:40:52PM +0800, Wizard wrote: >> Hi, Experts >> >> I am a newbie for linux kernel. I just read the code of >> find_next_system_ram() in kernel/resource.c >> >> I think the comment of this function is not exact. This says "Finds >> the lowest memory reosurce exists within [res->start.res->end)". >> >> While I think the code is to find the lowest memory resource overlaps >> [res->start, res->end). >> >> 308: if ((p->end >= start) && (p->start < end)) >> >> If I am not correct, please let me know. :) > > Right, > > hint for the future, do a "git annotate" on the file containing that > code - the patch adding the piece of code might (err, and should!) have > a verbose commit message explaining the situation. And it seems it does > have something to a degree, here's another hint: > > 58c1b5b079071 > > :-) > > But I agree that the comment over the function could use some more > verbosity on why the function needs to handle overlapping resources and > sections. Let's ask the author. > >> BTW, I find the "reosurce" is a typo. > > Yes, he could fix it while explaining the overlap :-). > > HTH. > > -- > Regards/Gruss, > Boris. >
-- Wizard
| |