linux-kernel - [PATCH 4.9 100/241] qed: Fix TM block ILT allocation

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 linux-cve-announce 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]

Message-Id: <20180319180755.343126670@linuxfoundation.org>
Date:   Mon, 19 Mar 2018 19:06:05 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org,
        Michal Kalderon <Michal.Kalderon@...ium.com>,
        Yuval Mintz <Yuval.Mintz@...ium.com>,
        "David S. Miller" <davem@...emloft.net>,
        Sasha Levin <alexander.levin@...rosoft.com>
Subject: [PATCH 4.9 100/241] qed: Fix TM block ILT allocation

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Michal Kalderon <Michal.Kalderon@...ium.com>


[ Upstream commit 44531ba45dbf3c23cc7ae0934ec9b33ef340ac56 ]

When configuring the HW timers block we should set the number of CIDs
up until the last CID that require timers, instead of only those CIDs
whose protocol needs timers support.

Today, the protocols that require HW timers' support have their CIDs
before any other protocol, but that would change in future [when we
add iWARP support].

Signed-off-by: Michal Kalderon <Michal.Kalderon@...ium.com>
Signed-off-by: Yuval Mintz <Yuval.Mintz@...ium.com>
Signed-off-by: David S. Miller <davem@...emloft.net>
Signed-off-by: Sasha Levin <alexander.levin@...rosoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
 drivers/net/ethernet/qlogic/qed/qed_cxt.c |   32 +++++++++++++++++++++++-------
 1 file changed, 25 insertions(+), 7 deletions(-)

--- a/drivers/net/ethernet/qlogic/qed/qed_cxt.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_cxt.c
@@ -271,16 +271,34 @@ struct qed_tm_iids {
 	u32 per_vf_tids;
 };
 
-static void qed_cxt_tm_iids(struct qed_cxt_mngr *p_mngr,
+static void qed_cxt_tm_iids(struct qed_hwfn *p_hwfn,
+			    struct qed_cxt_mngr *p_mngr,
 			    struct qed_tm_iids *iids)
 {
-	u32 i, j;
-
-	for (i = 0; i < MAX_CONN_TYPES; i++) {
+	bool tm_vf_required = false;
+	bool tm_required = false;
+	int i, j;
+
+	/* Timers is a special case -> we don't count how many cids require
+	 * timers but what's the max cid that will be used by the timer block.
+	 * therefore we traverse in reverse order, and once we hit a protocol
+	 * that requires the timers memory, we'll sum all the protocols up
+	 * to that one.
+	 */
+	for (i = MAX_CONN_TYPES - 1; i >= 0; i--) {
 		struct qed_conn_type_cfg *p_cfg = &p_mngr->conn_cfg[i];
 
-		if (tm_cid_proto(i)) {
+		if (tm_cid_proto(i) || tm_required) {
+			if (p_cfg->cid_count)
+				tm_required = true;
+
 			iids->pf_cids += p_cfg->cid_count;
+		}
+
+		if (tm_cid_proto(i) || tm_vf_required) {
+			if (p_cfg->cids_per_vf)
+				tm_vf_required = true;
+
 			iids->per_vf_cids += p_cfg->cids_per_vf;
 		}
 	}
@@ -696,7 +714,7 @@ int qed_cxt_cfg_ilt_compute(struct qed_h
 
 	/* TM PF */
 	p_cli = &p_mngr->clients[ILT_CLI_TM];
-	qed_cxt_tm_iids(p_mngr, &tm_iids);
+	qed_cxt_tm_iids(p_hwfn, p_mngr, &tm_iids);
 	total = tm_iids.pf_cids + tm_iids.pf_tids_total;
 	if (total) {
 		p_blk = &p_cli->pf_blks[0];
@@ -1591,7 +1609,7 @@ static void qed_tm_init_pf(struct qed_hw
 	u8 i;
 
 	memset(&tm_iids, 0, sizeof(tm_iids));
-	qed_cxt_tm_iids(p_mngr, &tm_iids);
+	qed_cxt_tm_iids(p_hwfn, p_mngr, &tm_iids);
 
 	/* @@@TBD No pre-scan for now */