Messages in this thread | | | Date | Tue, 22 Jul 2008 08:08:31 -0700 | From | Jeremy Fitzhardinge <> | Subject | Re: [git pull] x86 fixes |
| |
Johannes Weiner wrote: > Hi, > > Ingo Molnar <mingo@elte.hu> writes: > > >> Linus, >> >> Please pull the latest x86 fixes git tree from: >> >> git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git x86-fixes-for-linus >> >> Thanks, >> >> Ingo >> >> ------------------> >> Jan Kratochvil (1): >> x86: fix crash due to missing debugctlmsr on AMD K6-3 >> >> Jeremy Fitzhardinge (2): >> x86: rename PTE_MASK to PTE_PFN_MASK >> x86: add PTE_FLAGS_MASK >> > > Are you serious? It goes _this_ fast? The patch came into my mbox at > 8:00am this morning and you push it to Linus at 4:00pm already? > > What about the inconsistency it introduces? When I look at PAGE_MASK > for example, it masks out the PAGE offset. PTE_MASK masks out PTE > specifca from a value. > > Now, I assume PTE_PFN_MASK masks out the PFN. Oh, wait, it masks the > protection bits.
PAGE_MASK turns an address into its page address.
PTE_PFN_MASK takes a pte value and returns the pte's pfn portion (which is shifted so it's actually a page address).
In both cases, the X_MASK terminology means that X is extracted, not excluded. Which makes sense; if you have a packed bitfield containing multiple values, you wouldn't expect X to be the list of things *not* extracted.
(I would be happy to have a PAGE_OFFSET_MASK for all the places where you want to extract the offset.)
J
| |