netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: netfilter-devel@vger.kernel.org
Cc: kaber@trash.net, fw@strlen.de
Subject: [PATCH nft 4/7] evaluate: transfer right shifts to constant side
Date: Sat,  5 Dec 2015 20:04:23 +0100	[thread overview]
Message-ID: <1449342266-2756-4-git-send-email-pablo@netfilter.org> (raw)
In-Reply-To: <1449342266-2756-1-git-send-email-pablo@netfilter.org>

From: Patrick McHardy <kaber@trash.net>

This provides a generic way to transfer shifts from the left hand side
to the right hand constant side of a relational expression when
performing transformations from the evaluation step.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 src/evaluate.c | 30 ++++++++++++++++++++++++++----
 1 file changed, 26 insertions(+), 4 deletions(-)

diff --git a/src/evaluate.c b/src/evaluate.c
index 7aab6aa..d62870a 100644
--- a/src/evaluate.c
+++ b/src/evaluate.c
@@ -646,6 +646,7 @@ static int constant_binop_simplify(struct eval_ctx *ctx, struct expr **expr)
 		break;
 	case OP_LSHIFT:
 		assert(left->byteorder == BYTEORDER_HOST_ENDIAN);
+		mpz_set(val, left->value);
 		mpz_lshift_ui(val, mpz_get_uint32(right->value));
 		mpz_and(val, val, mask);
 		break;
@@ -1021,6 +1022,8 @@ static int binop_can_transfer(struct eval_ctx *ctx,
 			return expr_binary_error(ctx->msgs, right, left,
 						 "Comparison is always false");
 		return 1;
+	case OP_RSHIFT:
+		return 1;
 	case OP_XOR:
 		return 1;
 	default:
@@ -1038,6 +1041,10 @@ static int binop_transfer_one(struct eval_ctx *ctx,
 		(*right) = binop_expr_alloc(&(*right)->location, OP_RSHIFT,
 					    *right, expr_get(left->right));
 		break;
+	case OP_RSHIFT:
+		(*right) = binop_expr_alloc(&(*right)->location, OP_LSHIFT,
+					    *right, expr_get(left->right));
+		break;
 	case OP_XOR:
 		(*right) = binop_expr_alloc(&(*right)->location, OP_XOR,
 					    *right, expr_get(left->right));
@@ -1052,6 +1059,7 @@ static int binop_transfer_one(struct eval_ctx *ctx,
 static int binop_transfer(struct eval_ctx *ctx, struct expr **expr)
 {
 	struct expr *left = (*expr)->left, *i, *next;
+	unsigned int shift;
 	int err;
 
 	if (left->ops->type != EXPR_BINOP)
@@ -1083,10 +1091,24 @@ static int binop_transfer(struct eval_ctx *ctx, struct expr **expr)
 		return 0;
 	}
 
-	left = expr_get((*expr)->left->left);
-	left->dtype = (*expr)->left->dtype;
-	expr_free((*expr)->left);
-	(*expr)->left = left;
+	switch (left->op) {
+	case OP_RSHIFT:
+		/* Mask out the bits the shift would have masked out */
+		shift = mpz_get_uint8(left->right->value);
+		mpz_bitmask(left->right->value, left->left->len);
+		mpz_lshift_ui(left->right->value, shift);
+		left->op = OP_AND;
+		break;
+	case OP_LSHIFT:
+	case OP_XOR:
+		left = expr_get((*expr)->left->left);
+		left->dtype = (*expr)->left->dtype;
+		expr_free((*expr)->left);
+		(*expr)->left = left;
+		break;
+	default:
+		BUG("invalid binop operation %u", left->op);
+	}
 	return 0;
 }
 
-- 
2.1.4


  parent reply	other threads:[~2015-12-05 19:04 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-12-05 19:04 [PATCH nft 1/7,v2] tests: vlan pcp and cfi are located in the first byte Pablo Neira Ayuso
2015-12-05 19:04 ` [PATCH nft 2/7,v2] src: fix sub-byte protocol header definitions Pablo Neira Ayuso
2015-12-14 17:38   ` Pablo Neira Ayuso
2015-12-05 19:04 ` [PATCH nft 3/7] netlink_delinearize: postprocess expression before range merge Pablo Neira Ayuso
2015-12-05 19:04 ` Pablo Neira Ayuso [this message]
2015-12-05 19:04 ` [PATCH nft 5/7] evaluate: transfer right shifts to range side Pablo Neira Ayuso
2015-12-05 19:04 ` [PATCH nft 6/7] evaluate: transfer right shifts to set reference side Pablo Neira Ayuso
2015-12-05 19:04 ` [PATCH nft 7/7] src: move payload sub-byte matching to the evaluation step Pablo Neira Ayuso

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1449342266-2756-4-git-send-email-pablo@netfilter.org \
    --to=pablo@netfilter.org \
    --cc=fw@strlen.de \
    --cc=kaber@trash.net \
    --cc=netfilter-devel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).