Create manpage for pmap_quick_enter_page() and pmap_quick_remove_page()
Details
- Reviewers
kib wblock brueffer - Group Reviewers
manpages - Commits
- rS286411: Create man page for pmap_quick_enter_page(9) and pmap_quick_remove_page(9)
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
share/man/man9/pmap_quick_enter_page.9 | ||
---|---|---|
34 ↗ | (On Diff #7697) | address space |
46 ↗ | (On Diff #7697) | I do not think that wired is any sort of requirement for the state of the page m. Caller must ensure that the page is not reused for something undesirable, but thats all. E.g., the page could be unmanaged, or it could be managed but held instead of wired, or it could be busy. Any of the listed condition are enough to keep the m content stable. |
56 ↗ | (On Diff #7697) | s/.$/, making the KVA frame used by pmap_quick_enter_page() available for reuse./ Or, some other way to say that the calling thread must not recurse into the dynamic region marked by pmap_quick_enter/remove. |
62 ↗ | (On Diff #7697) | This sentence does not describe implementation, it should be moved to the main body of the function description. In particular, main text must say that the region cannot be nested, that the locking cannot be used in the region, and remove must be called as soon as possible. Also, the context where the function can be called should be mentioned (thread but not interrupt handler, no spinlock may be owned when called). |
72 ↗ | (On Diff #7697) | See above, this should be moved from notes. |
77 ↗ | (On Diff #7697) | 'Cannot fail' claim must be also moved. |
79 ↗ | (On Diff #7697) | .... |
83 ↗ | (On Diff #7697) | No. |
92 ↗ | (On Diff #7697) | I propose not to say this in the man page. |
Overall, it is fine WRT the content.
Note that the each sentence in the mdoc source should start on the new line.
share/man/man9/pmap_quick_enter_page.9 | ||
---|---|---|
86 ↗ | (On Diff #7713) | Note that even if the page is swapped, nothing wrong would happen with the mapping. Both the page and the *m are still there. It is up to the caller to decide is it worrying. |
share/man/man9/pmap_quick_enter_page.9 | ||
---|---|---|
57 ↗ | (On Diff #7720) | The "at" should be on a separate line: .Fn pmap_quick_enter_page at |
80 ↗ | (On Diff #7720) | "cannot fail" implies that they always work, but the next two sentences have some exceptions. |
86 ↗ | (On Diff #7720) | "is valid" is ambiguous for the reader, who might not know what that means. Maybe "in place" or "active" or something like that? |
The man page changes look good to me. Before commit, please check with igor -R in addition to the mandoc -Tlint that brueffer mentioned. Thanks!