x86_64: mm: support numa migration pte entry #379
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The PRESENT bit alone does not indicate whether a page actually exists. The PRESENT bit is also cleared for numa migration pte entries, even though the page is present. Therefore, to determine if a page truly exists, we need to check pte and convert it to phys in the same way the kernel does. Signed-off-by: Hao Li <[email protected]>
osandov
requested changes
Jan 4, 2024
| table = entry & ADDRESS_MASK; | ||
| if (!(entry & PRESENT) || (entry & PSE) || level == 0) { | ||
| table = entry & PTE_PFN_MASK; | ||
| if (!(entry & _PAGE_PRESENT) || (entry & PSE) || level == 0) { |
There was a problem hiding this comment.
Can huge pages also have numa migration entries? pmd_present() in the kernel checks _PAGE_PROTNONE, so maybe 2MB huge pages can? If so, this check probably needs to be slightly different (and depend on the level).
Comment on lines
+663
to
+664
| if (pte_present(entry)) | ||
| *phys_addr_ret = pte_phys(entry); |
There was a problem hiding this comment.
This will also get entries for higher levels (for huge pages or unmapped levels), so the pte_* names are not completely accurate. The functions themselves probably also need to depend on the level as noted above.
|
|
||
| static inline uint64_t protnone_mask(uint64_t val) | ||
| { | ||
| return __pte_needs_invert(val) ? ~0ull : 0; |
There was a problem hiding this comment.
In theory, ~0ull could be a different size than uint64_t, so I'd feel better about:
Suggested change
| return __pte_needs_invert(val) ? ~0ull : 0; | |
| return __pte_needs_invert(val) ? -1 : 0; |
Comment on lines
+585
to
+598
| static inline bool __pte_needs_invert(uint64_t val) | ||
| { | ||
| return val && !(val & _PAGE_PRESENT); | ||
| } | ||
|
|
||
| static inline uint64_t protnone_mask(uint64_t val) | ||
| { | ||
| return __pte_needs_invert(val) ? ~0ull : 0; | ||
| } | ||
|
|
||
| static inline uint64_t pte_flags(uint64_t pte) | ||
| { | ||
| return pte & ((1 << PAGE_SHIFT) - 1); | ||
| } |
There was a problem hiding this comment.
I'd prefer to fold these small helpers into their callers.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The PRESENT bit alone does not indicate whether a page actually exists. The PRESENT bit is also cleared for numa migration pte entries, even though the page is present. Therefore, to determine if a page truly exists, we need to check pte and convert it to phys in the same way the kernel does.
Fix #370
Signed-off-by: Hao Li [email protected]