Hi Stan,
(For that matter, let me say "I'm planning to use it in the next few milliseconds... please fetch it into memory behind my back now, while I do other things").
This one part of your wishes may be possible, but I'm not sure for what proportion of memory, with madvise(..., MADV_WILLNEED). Regards, James. On Sat, Mar 24, 2012 at 8:00 AM, <darwin-kernel-request@lists.apple.com> wrote: <snip>
------------------------------
Message: 9 Date: Fri, 23 Mar 2012 10:44:42 -0700 From: Stan Sieler <sieler@me.com> To: "Gerriet M. Denkmann" <gerriet@mdenkmann.de> Cc: "darwin-kernel@lists.apple.com Kernel" <darwin-kernel@lists.apple.com> Subject: Re: vmmap Message-ID: <9772AEC8-AC3B-4746-B3A2-0EFC1943B3CD@me.com> Content-Type: text/plain; CHARSET=US-ASCII
Re:
Is the source of the Unix tool vmmap available somewhere? Also, are you trying to get this information for debugging purposes, or to make some meaningful runtime decision on an app that you ship to real users? First I want this information for my own entertainment and to educate myself. Publishing the resulting app for real users might be something to be thought about in the far future.
I can sympathize with Gerriet ... I asked similar questions about HP's MPE/iX operating system, and that's what led me to be able to write tools that were eventually products used by users *and* HP.
(With the HP 3000, we had a strong users group ... which helped provide much more visibility into the kernel and much more interaction with the kernel lab.)
I find memory (real and virtual) to be one of the hardest things to get good information about on operating systems other than the now-discontinued MPE/iX. I like asking questions: is this virtual address valid? Is it readable by me? Writable? Is this virtual page in memory? Is it dirty? Was it referenced recently? Is it frozen in memory? (Let me freeze, and later unfreeze it.) (For that matter, let me say "I'm planning to use it in the next few milliseconds... please fetch it into memory behind my back now, while I do other things").
These are all reasonable questions to ask ... one can easily see how they can lead to better/faster code in some cases.
Stan
_______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel@lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.... This email sent to site_archiver@lists.apple.com
participants (1)
-
James C