nftables/backport-evaluate-fix-shift-exponent-underflow-in-concatenation-evaluation.patch
zhang-hao-jon dba9b0e264 nftables:fix some pathces from commity
(cherry picked from commit 3cf66f5ae827220d1b43d0628749805eb9ed8133)
2023-04-06 15:08:38 +08:00

40 lines
1.2 KiB
Diff

From 0fe79458cb5ae36d838f0e5a5dc5cc6f332cac03 Mon Sep 17 00:00:00 2001
From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Thu, 22 Dec 2022 11:23:00 +0100
Subject: [PATCH] evaluate: fix shift exponent underflow in concatenation
evaluation
There is an underflow of the index that iterates over the concatenation:
../include/datatype.h:292:15: runtime error: shift exponent 4294967290 is too large for 32-bit type 'unsigned int'
set the datatype to invalid which is fine to evaluate a concatenation
in a set/map statement.
Conflict: NA
Reference: https://git.netfilter.org/nftables/commit?id=0fe79458cb5ae36d838f0e5a5dc5cc6f332cac03
Update b8e1940aa190 ("tests: add a test case for map update from packet
path with concat") so it does not need a workaround to work.
---
src/evaluate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/evaluate.c b/src/evaluate.c
index 37d0bff..36ce38b 100644
--- a/src/evaluate.c
+++ b/src/evaluate.c
@@ -1265,7 +1265,7 @@ static int expr_evaluate_concat(struct eval_ctx *ctx, struct expr **expr)
"expecting %s",
dtype->desc);
- if (dtype == NULL)
+ if (dtype == NULL || off == 0)
tmp = datatype_lookup(TYPE_INVALID);
else
tmp = concat_subtype_lookup(type, --off);
--
2.33.0