Thread (4 messages) flat view 4 messages, 2 authors, 15d ago

Re: [PATCH iwl-net] ice: fix bound parser hash offset before reading packet data

From: Simon Horman <horms@kernel.org>
Date: 2026-08-24 11:08:47
Also in: intel-wired-lan

On Thu, Aug 20, 2026 at 01:43:14PM +0200, Aleksandr Loktionov wrote:
quoted hunk ↗ jump to hunk
ice_rt_ho_set() uses the HO register as the starting offset of an
ICE_GPR_HV_SIZE-byte memcpy() out of rt->pkt_buf. Potentially HO can
be advanced by user-controlled data reachable through
ice_parse_raw_rss_pattern() -> ice_parser_run() ->
ice_parser_rt_execute() -> ice_rt_gpr_set() -> ice_rt_ho_set(), i.e. a
VF-supplied raw RSS pattern (virt/rss.c), with no bound against the
size of pkt_buf.

Clamp HO to the last offset from which ICE_GPR_HV_SIZE bytes can still
be read out of pkt_buf, deriving the limit from sizeof(rt->pkt_buf)
so it stays correct if the packet buffer layout changes.

Fixes: 9a4c07aaa0f5 ("ice: add parser execution main loop")
Cc: stable@vger.kernel.org
Signed-off-by: Aleksandr Loktionov <redacted>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_parser_rt.c | 2 ++
 1 file changed, 2 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_parser_rt.c b/drivers/net/ethernet/intel/ice/ice_parser_rt.c
index 3995d66..bfdb50b 100644
--- a/drivers/net/ethernet/intel/ice/ice_parser_rt.c
+++ b/drivers/net/ethernet/intel/ice/ice_parser_rt.c
@@ -10,6 +10,8 @@ static void ice_rt_tsr_set(struct ice_parser_rt *rt, u16 tsr)
 
 static void ice_rt_ho_set(struct ice_parser_rt *rt, u16 ho)
 {
+	/* keep the ICE_GPR_HV_SIZE-byte read below within pkt_buf */
+	ho = min_t(u16, ho, sizeof(rt->pkt_buf) - ICE_GPR_HV_SIZE);
 	rt->gpr[ICE_GPR_HO_IDX] = ho;
 	memcpy(&rt->gpr[ICE_GPR_HV_IDX], &rt->pkt_buf[ho], ICE_GPR_HV_SIZE);
 }
Hi,

There is an AI-generated review of this patch available at
https://sashiko.dev/#/patchset/20260820114314.646344-1-aleksandr.loktionov%40intel.com

The first item flagged there seems of concern
and I am wondering if you could respond to it.

  Does this new clamp introduce an infinite loop?

  In ice_parser_rt_execute(), the loop termination condition is:

      if (rt->gpr[ICE_GPR_HO_IDX] >= rt->pkt_len) {
          break;
      }

  If a guest VM VF submits a raw RSS pattern with a pkt_len greater than 504
  bytes, ice_parser_rt_pktbuf_set() saves that unclamped length directly into
  rt->pkt_len (e.g., 1024).

  With this patch, the HO register is capped at 488 (sizeof(rt->pkt_buf) -
  ICE_GPR_HV_SIZE). Since 488 will never be greater than or equal to 1024,
  can this cause the parser to loop indefinitely if the state machine
  processes a repetitive sequence?
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help