__vmap_pages_range_noflush() and friends can install some PTEs before failing and leave them mapped, and the kernel callers do not agree on who cleans them up. pcpu_map_pages() and kmsan_ioremap_page_range() roll back what they mapped before the failure, while vm_module_tags_populate() and the __GFP_NOFAIL retry loop in __vmalloc_area_node() rely on the mapping functions cleaning up and do not call anything like vunmap_range() themselves. When the same range is mapped again, the attempt hits the leftovers and fails, with a BUG() in vmap_pte_range() for huge mappings and a warning on the small-page path.
After discussing with Suren and Ulad [1], we move the rollback into the mapping functions. A failed vmap leaves nothing mapped, and the callers no longer clean up after it. Patch 1 is the KMSAN part. The shadow and the origin of a range are mapped by two separate calls, so the rollback there has to cover both. Patch 2 is its disabled-KMSAN counterpart. Patches 3 to 7 clean up the range checks and error reporting: unaligned ranges are clamped or rejected, the callers get the real error code instead of a flat -ENOMEM, and the cache is only flushed once the mapping succeeded. Patch 8 adds the rollback to __vmap_pages_range_noflush() and vmap_page_range(). Patches 9 and 10 drop the powerpc and percpu cleanups that are not needed anymore. Changes since v2: - new patches 3 to 7 fix the range checks and the error reporting; the empty range check of v2's patch 3 moved into patch 6 - patch 8 (was patch 3) skips cond_resched() in the rollback when the caller's gfp mask cannot block - commit messages reworded [1] https://lore.kernel.org/all/[email protected]/ v2 link: https://lore.kernel.org/all/[email protected]/ v1 link: https://lore.kernel.org/all/[email protected]/ Hao Ge (10): mm/kmsan: undo the shadow mapping when the origin mapping fails mm/kmsan: skip vunmap_range_noflush when KMSAN is disabled mm/vmalloc: clamp an unaligned range in vunmap_pte_range() mm/vmalloc: propagate errors in vmap_pages_p*d_range() mm/vmalloc: reject an unaligned range in vmap_pages_pte_range() mm/vmalloc: reject empty and unaligned ranges in __vmap_pages_range_noflush() mm/vmalloc: only flush the cache after the vmap succeeded mm/vmalloc: undo partial mappings inside the mapping functions powerpc: drop redundant unmaps of failed vmap mappings mm/percpu: stop unmapping the CPU that failed to map arch/powerpc/kernel/isa-bridge.c | 5 +- arch/powerpc/kernel/pci_64.c | 4 +- include/linux/kmsan.h | 8 +- mm/kmsan/hooks.c | 21 ++++-- mm/kmsan/shadow.c | 9 ++- mm/percpu-vm.c | 5 +- mm/vmalloc.c | 125 ++++++++++++++++++++++--------- mm/vmalloc.h | 6 +- 8 files changed, 127 insertions(+), 56 deletions(-) -- 2.25.1
