Skip to contentSkip to conversation

ntb_hw_switchtec: Fix shift-out-of-bounds in switchtec_ntb_mw_set_trans

PATCH In review Yajun Deng 3 participants 3 messages Aug 16, 2023 – Aug 17, 2023

Subsystems (3)
lore
Share & tools
lore

Yajun Deng

author
#0 lore raw

PATCH 1/1 ntb_hw_switchtec: Fix shift-out-of-bounds in switchtec_ntb_mw_set_trans

There is a kernel API ntb_mw_clear_trans() would pass 0 to both addr and size. This would make xlate_pos negative.

[   23.734156] switchtec switchtec0: MW 0: part 0 addr 0x0000000000000000 size 0x0000000000000000
[   23.734158] ================================================================================
[   23.734172] UBSAN: shift-out-of-bounds in drivers/ntb/hw/mscc/ntb_hw_switchtec.c:293:7
[   23.734418] shift exponent -1 is negative

Ensuring xlate_pos is a positive or zero before BIT.

Fixes 1e2fd202f859 ("ntb_hw_switchtec: Check for alignment of the buffer in mw_set_trans()")
---
 drivers/ntb/hw/mscc/ntb_hw_switchtec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/ntb/hw/mscc/ntb_hw_switchtec.c +1 -1
index d6bbcc7b5b90..21468d4fef64 100644
@@ -288,7 +288,7 @@ static int switchtec_ntb_mw_set_trans(struct ntb_dev *ntb, int pidx, int widx,
 	if (size != 0 && xlate_pos < 12)
 		return -EINVAL;
 
-	if (!IS_ALIGNED(addr, BIT_ULL(xlate_pos))) {
+	if (xlate_pos >= 0 && !IS_ALIGNED(addr, BIT_ULL(xlate_pos))) {
 		/*
 		 * In certain circumstances we can get a buffer that is
 		 * not aligned to its size. (Most of the time
Signature · 1 line
2.25.1
maintainer Reviewed-by on v1 1/1
#1 lore raw
Quoted 8 lines · Yajun Deng

There is a kernel API ntb_mw_clear_trans() would pass 0 to both addr and size. This would make xlate_pos negative.

[   23.734156] switchtec switchtec0: MW 0: part 0 addr 0x0000000000000000 size 0x0000000000000000
[   23.734158] ================================================================================
[   23.734172] UBSAN: shift-out-of-bounds in drivers/ntb/hw/mscc/ntb_hw_switchtec.c:293:7
[   23.734418] shift exponent -1 is negative
Quoted 4 lines

Ensuring xlate_pos is a positive or zero before BIT.

Fixes 1e2fd202f859 ("ntb_hw_switchtec: Check for alignment of the buffer in mw_set_trans()")

That makes sense. Thanks!

Logan

#2 lore raw
In reply to Yajun Deng
Quoted 9 lines · Yajun Deng

There is a kernel API ntb_mw_clear_trans() would pass 0 to both addr and size. This would make xlate_pos negative.

[   23.734156] switchtec switchtec0: MW 0: part 0 addr 0x0000000000000000 size 0x0000000000000000
[   23.734158] ================================================================================
[   23.734172] UBSAN: shift-out-of-bounds in drivers/ntb/hw/mscc/ntb_hw_switchtec.c:293:7
[   23.734418] shift exponent -1 is negative

Ensuring xlate_pos is a positive or zero before BIT.

I assume Kurt or Logan will apply this and no need to repost for this, but if you do repost for some reason, the timestamps and separator lines above are clutter and don't contribute to understanding the problem.

Also s/Ensuring/Ensure/

Quoted 22 lines
Fixes 1e2fd202f859 ("ntb_hw_switchtec: Check for alignment of the buffer in mw_set_trans()")
---
 drivers/ntb/hw/mscc/ntb_hw_switchtec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/ntb/hw/mscc/ntb_hw_switchtec.c +1 -1
index d6bbcc7b5b90..21468d4fef64 100644
@@ -288,7 +288,7 @@ static int switchtec_ntb_mw_set_trans(struct ntb_dev *ntb, int pidx, int widx,
 	if (size != 0 && xlate_pos < 12)
 		return -EINVAL;
 
-	if (!IS_ALIGNED(addr, BIT_ULL(xlate_pos))) {
+	if (xlate_pos >= 0 && !IS_ALIGNED(addr, BIT_ULL(xlate_pos))) {
 		/*
 		 * In certain circumstances we can get a buffer that is
 		 * not aligned to its size. (Most of the time
Signature · 1 line
2.25.1