netdev - [RFCv4 PATCH net-next 2/3] net-next: ethtool: move checks before rtnl_lock() in ethnl_set

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 31 Mar 2022 16:43:41 +0800
From:   Jie Wang <wangjie125@...wei.com>
To:     <mkubecek@...e.cz>, <davem@...emloft.net>, <kuba@...nel.org>,
        <wangjie125@...wei.com>
CC:     <netdev@...r.kernel.org>, <huangguangbin2@...wei.com>,
        <lipeng321@...wei.com>, <shenjian15@...wei.com>,
        <moyufeng@...wei.com>, <linyunsheng@...wei.com>,
        <salil.mehta@...wei.com>, <chenhao288@...ilicon.com>
Subject: [RFCv4 PATCH net-next 2/3] net-next: ethtool: move checks before rtnl_lock() in ethnl_set_rings

Currently these two checks in ethnl_set_rings are added after rtnl_lock()
which will do useless works if the request is invalid.

So this patch moves these checks before the rtnl_lock() to avoid these
costs.

Signed-off-by: Jie Wang <wangjie125@...wei.com>
---
 net/ethtool/rings.c | 38 ++++++++++++++++++++------------------
 1 file changed, 20 insertions(+), 18 deletions(-)

diff --git a/net/ethtool/rings.c b/net/ethtool/rings.c
index 9ed60c507d97..46415d8fc256 100644
--- a/net/ethtool/rings.c
+++ b/net/ethtool/rings.c
@@ -152,6 +152,26 @@ int ethnl_set_rings(struct sk_buff *skb, struct genl_info *info)
 	if (!ops->get_ringparam || !ops->set_ringparam)
 		goto out_dev;
 
+	if (tb[ETHTOOL_A_RINGS_RX_BUF_LEN] &&
+	    nla_get_u32(tb[ETHTOOL_A_RINGS_RX_BUF_LEN]) != 0 &&
+	    !(ops->supported_ring_params & ETHTOOL_RING_USE_RX_BUF_LEN)) {
+		ret = -EOPNOTSUPP;
+		NL_SET_ERR_MSG_ATTR(info->extack,
+				    tb[ETHTOOL_A_RINGS_RX_BUF_LEN],
+				    "setting rx buf len not supported");
+		goto out_dev;
+	}
+
+	if (tb[ETHTOOL_A_RINGS_CQE_SIZE] &&
+	    nla_get_u32(tb[ETHTOOL_A_RINGS_CQE_SIZE]) &&
+	    !(ops->supported_ring_params & ETHTOOL_RING_USE_CQE_SIZE)) {
+		ret = -EOPNOTSUPP;
+		NL_SET_ERR_MSG_ATTR(info->extack,
+				    tb[ETHTOOL_A_RINGS_CQE_SIZE],
+				    "setting cqe size not supported");
+		goto out_dev;
+	}
+
 	if (tb[ETHTOOL_A_RINGS_TX_PUSH] &&
 	    !(ops->supported_ring_params & ETHTOOL_RING_USE_TX_PUSH)) {
 		ret = -EOPNOTSUPP;
@@ -201,24 +221,6 @@ int ethnl_set_rings(struct sk_buff *skb, struct genl_info *info)
 		goto out_ops;
 	}
 
-	if (kernel_ringparam.rx_buf_len != 0 &&
-	    !(ops->supported_ring_params & ETHTOOL_RING_USE_RX_BUF_LEN)) {
-		ret = -EOPNOTSUPP;
-		NL_SET_ERR_MSG_ATTR(info->extack,
-				    tb[ETHTOOL_A_RINGS_RX_BUF_LEN],
-				    "setting rx buf len not supported");
-		goto out_ops;
-	}
-
-	if (kernel_ringparam.cqe_size &&
-	    !(ops->supported_ring_params & ETHTOOL_RING_USE_CQE_SIZE)) {
-		ret = -EOPNOTSUPP;
-		NL_SET_ERR_MSG_ATTR(info->extack,
-				    tb[ETHTOOL_A_RINGS_CQE_SIZE],
-				    "setting cqe size not supported");
-		goto out_ops;
-	}
-
 	ret = dev->ethtool_ops->set_ringparam(dev, &ringparam,
 					      &kernel_ringparam, info->extack);
 	if (ret < 0)
-- 
2.33.0