cxl/region: Use do_div() for 64-bit modulo operation
Alison Schofield
maintainerPATCH 1/1 cxl/region: Use do_div() for 64-bit modulo operation
div64_u64_rem() was the wrong choice for doing a modulo operation and it was used incorrectly, causing a kernel oops by passing NULL as the remainder parameter. Replace it with the do_div() helper that does the intended math (gran_offset = offset % gran) and is architecture safe.
This bug appeared during testing of unaligned address translations. The visibility to userspace would be limited to folks doing poison injection or clear by HPA on unaligned regions.
--- drivers/cxl/core/region.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/cxl/core/region.c +2 -1
@@ -3324,6 +3324,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr,u64 interleave_width, interleave_index;u64 gran, gran_offset, dpa_offset;u64 hpa = p->res->start + offset; +u64 tmp = offset;/* * Unaligned addresses are not algebraically invertible. Calculate
@@ -3333,7 +3334,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr, gran = cxld->interleave_granularity; interleave_width = gran * cxld->interleave_ways; interleave_index =div64_u64 (offset, interleave_width); - gran_offset =div64_u64_rem (offset, gran,NULL ); + gran_offset =do_div (tmp, gran); dpa_offset = interleave_index * gran + gran_offset;
base-commit: 78b50b5984629d362f826e9615ce4599f429716e
Signature · 1 line
2.37.3
Quoted 6 lines · Alison Schofield
div64_u64_rem()was the wrong choice for doing a modulo operation and it was used incorrectly, causing a kernel oops by passing NULL as the remainder parameter. Replace it with thedo_div()helper that does the intended math (gran_offset= offset % gran) and is architecture safe.
Quoted 6 lines
This bug appeared during testing of unaligned address translations. The visibility to userspace would be limited to folks doing poison injection or clear by HPA on unaligned regions.
Fixes 78b50b598462 ("cxl/region: Translate HPA to DPA and memdev in unaligned regions")Signed-off-by Alison Schofield <[email protected]>
Thanks Alison. I'll get this applied to next
Quoted 27 lines
--- drivers/cxl/core/region.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)drivers/cxl/core/region.c +2
-1@@ -3324,6 +3324,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr,u64 interleave_width, interleave_index;u64 gran, gran_offset, dpa_offset;u64 hpa = p->res->start + offset; +u64 tmp = offset;/* * Unaligned addresses are not algebraically invertible. Calculate @@ -3333,7 +3334,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr, gran = cxld->interleave_granularity; interleave_width = gran * cxld->interleave_ways; interleave_index =div64_u64 (offset, interleave_width); - gran_offset =div64_u64_rem (offset, gran,NULL ); + gran_offset =do_div (tmp, gran); dpa_offset = interleave_index * gran + gran_offset;base-commit:
78b50b5984629d362f826e9615ce4599f429716e
Quoted 6 lines · Alison Schofield
div64_u64_rem()was the wrong choice for doing a modulo operation and it was used incorrectly, causing a kernel oops by passing NULL as the remainder parameter. Replace it with thedo_div()helper that does the intended math (gran_offset= offset % gran) and is architecture safe.
Quoted 6 lines
This bug appeared during testing of unaligned address translations. The visibility to userspace would be limited to folks doing poison injection or clear by HPA on unaligned regions.
Fixes 78b50b598462 ("cxl/region: Translate HPA to DPA and memdev in unaligned regions")Signed-off-by Alison Schofield <[email protected]>
Applied to cxl/next345f23df5d08fe58fb865cd17b696cd345455224
Quoted 27 lines
--- drivers/cxl/core/region.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)drivers/cxl/core/region.c +2
-1@@ -3324,6 +3324,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr,u64 interleave_width, interleave_index;u64 gran, gran_offset, dpa_offset;u64 hpa = p->res->start + offset; +u64 tmp = offset;/* * Unaligned addresses are not algebraically invertible. Calculate @@ -3333,7 +3334,7 @@ static int unaligned_region_offset_to_dpa_result(struct cxl_region *cxlr, gran = cxld->interleave_granularity; interleave_width = gran * cxld->interleave_ways; interleave_index =div64_u64 (offset, interleave_width); - gran_offset =div64_u64_rem (offset, gran,NULL ); + gran_offset =do_div (tmp, gran); dpa_offset = interleave_index * gran + gran_offset;base-commit:
78b50b5984629d362f826e9615ce4599f429716e