linux-kernel - [PATCH 5.4 350/434] ice: Only disable VF state when freeing each VF resources

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: <20191229172725.220428240@linuxfoundation.org>
Date:   Sun, 29 Dec 2019 18:26:43 +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,
        Akeem G Abodunrin <akeem.g.abodunrin@...el.com>,
        Andrew Bowers <andrewx.bowers@...el.com>,
        Jeff Kirsher <jeffrey.t.kirsher@...el.com>,
        Sasha Levin <sashal@...nel.org>
Subject: [PATCH 5.4 350/434] ice: Only disable VF state when freeing each VF resources

From: Akeem G Abodunrin <akeem.g.abodunrin@...el.com>

[ Upstream commit 1f9639d2fb9188a59acafae9dea626391c442a8d ]

It is wrong to set PF disable state flag for all VFs when freeing VF
resources - Instead, we should set VF disable state flag for each VF with
its resources being returned to the device. Right now, all VF opcodes,
mailbox communication to clear its resources as well fails - since we
already indicate that PF is in disable state, with all VFs not active. In
addition, we don't need to notify VF that PF is intending to reset it, if
it is already in disabled state.

Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@...el.com>
Tested-by: Andrew Bowers <andrewx.bowers@...el.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@...el.com>
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
 drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index b45797f39b2f..c0637a0cbfe8 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -317,8 +317,9 @@ void ice_free_vfs(struct ice_pf *pf)
 	pf->num_alloc_vfs = 0;
 	for (i = 0; i < tmp; i++) {
 		if (test_bit(ICE_VF_STATE_INIT, pf->vf[i].vf_states)) {
-			/* disable VF qp mappings */
+			/* disable VF qp mappings and set VF disable state */
 			ice_dis_vf_mappings(&pf->vf[i]);
+			set_bit(ICE_VF_STATE_DIS, pf->vf[i].vf_states);
 			ice_free_vf_res(&pf->vf[i]);
 		}
 	}
@@ -1287,9 +1288,12 @@ static void ice_vc_notify_vf_reset(struct ice_vf *vf)
 	if (!vf || vf->vf_id >= vf->pf->num_alloc_vfs)
 		return;
 
-	/* verify if the VF is in either init or active before proceeding */
-	if (!test_bit(ICE_VF_STATE_INIT, vf->vf_states) &&
-	    !test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states))
+	/* Bail out if VF is in disabled state, neither initialized, nor active
+	 * state - otherwise proceed with notifications
+	 */
+	if ((!test_bit(ICE_VF_STATE_INIT, vf->vf_states) &&
+	     !test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states)) ||
+	    test_bit(ICE_VF_STATE_DIS, vf->vf_states))
 		return;
 
 	pfe.event = VIRTCHNL_EVENT_RESET_IMPENDING;
-- 
2.20.1