diff options
author | Jason Gunthorpe <jgg@nvidia.com> | 2022-12-07 23:44:42 +0300 |
---|---|---|
committer | Jason Gunthorpe <jgg@nvidia.com> | 2022-12-09 22:20:37 +0300 |
commit | a26fa392068d1dcdf781397b7a7dd908dd68f030 (patch) | |
tree | 728a1b102e8b808e82cdf331995826f5c948912e /drivers | |
parent | c9b8a83a8f2dca9f82288a621595a6a5970cdc5e (diff) | |
download | linux-a26fa392068d1dcdf781397b7a7dd908dd68f030.tar.xz |
iommufd: Improve a few unclear bits of code
Correct a few items noticed late in review:
- We should assert that the math in batch_clear_carry() doesn't underflow
- user->locked should be -1 not 0 sicne we just did mmput
- npages should not have been recalculated, it already has that value
No functional change.
Fixes: 8d160cd4d506 ("iommufd: Algorithms for PFN storage")
Link: https://lore.kernel.org/r/2-v1-0362a1a1c034+98-iommufd_fixes1_jgg@nvidia.com
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reported-by: Binbin Wu <binbin.wu@linux.intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Diffstat (limited to 'drivers')
-rw-r--r-- | drivers/iommu/iommufd/pages.c | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/drivers/iommu/iommufd/pages.c b/drivers/iommu/iommufd/pages.c index fccdba782cb6..c77177229648 100644 --- a/drivers/iommu/iommufd/pages.c +++ b/drivers/iommu/iommufd/pages.c @@ -289,6 +289,10 @@ static void batch_clear_carry(struct pfn_batch *batch, unsigned int keep_pfns) if (!keep_pfns) return batch_clear(batch); + if (IS_ENABLED(CONFIG_IOMMUFD_TEST)) + WARN_ON(!batch->end || + batch->npfns[batch->end - 1] < keep_pfns); + batch->total_pfns = keep_pfns; batch->npfns[0] = keep_pfns; batch->pfns[0] = batch->pfns[batch->end - 1] + @@ -723,7 +727,7 @@ static void pfn_reader_user_destroy(struct pfn_reader_user *user, mmap_read_unlock(pages->source_mm); if (pages->source_mm != current->mm) mmput(pages->source_mm); - user->locked = 0; + user->locked = -1; } kfree(user->upages); @@ -810,7 +814,6 @@ static int incr_user_locked_vm(struct iopt_pages *pages, unsigned long npages) lock_limit = task_rlimit(pages->source_task, RLIMIT_MEMLOCK) >> PAGE_SHIFT; - npages = pages->npinned - pages->last_npinned; do { cur_pages = atomic_long_read(&pages->source_user->locked_vm); new_pages = cur_pages + npages; |