lkml.org 
[lkml]   [2022]   [Mar]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH 5.10 056/105] ice: fix concurrent reset and removal of VFs
Date
From: Jacob Keller <jacob.e.keller@intel.com>

commit fadead80fe4c033b5e514fcbadd20b55c4494112 upstream.

Commit c503e63200c6 ("ice: Stop processing VF messages during teardown")
introduced a driver state flag, ICE_VF_DEINIT_IN_PROGRESS, which is
intended to prevent some issues with concurrently handling messages from
VFs while tearing down the VFs.

This change was motivated by crashes caused while tearing down and
bringing up VFs in rapid succession.

It turns out that the fix actually introduces issues with the VF driver
caused because the PF no longer responds to any messages sent by the VF
during its .remove routine. This results in the VF potentially removing
its DMA memory before the PF has shut down the device queues.

Additionally, the fix doesn't actually resolve concurrency issues within
the ice driver. It is possible for a VF to initiate a reset just prior
to the ice driver removing VFs. This can result in the remove task
concurrently operating while the VF is being reset. This results in
similar memory corruption and panics purportedly fixed by that commit.

Fix this concurrency at its root by protecting both the reset and
removal flows using the existing VF cfg_lock. This ensures that we
cannot remove the VF while any outstanding critical tasks such as a
virtchnl message or a reset are occurring.

This locking change also fixes the root cause originally fixed by commit
c503e63200c6 ("ice: Stop processing VF messages during teardown"), so we
can simply revert it.

Note that I kept these two changes together because simply reverting the
original commit alone would leave the driver vulnerable to worse race
conditions.

Fixes: c503e63200c6 ("ice: Stop processing VF messages during teardown")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/net/ethernet/intel/ice/ice_main.c | 2 +
drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 35 ++++++++++++++++-------
2 files changed, 27 insertions(+), 10 deletions(-)

--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -1602,7 +1602,9 @@ static void ice_handle_mdd_event(struct
* reset, so print the event prior to reset.
*/
ice_print_vf_rx_mdd_event(vf);
+ mutex_lock(&pf->vf[i].cfg_lock);
ice_reset_vf(&pf->vf[i], false);
+ mutex_unlock(&pf->vf[i].cfg_lock);
}
}
}
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -360,22 +360,26 @@ void ice_free_vfs(struct ice_pf *pf)
else
dev_warn(dev, "VFs are assigned - not disabling SR-IOV\n");

- /* Avoid wait time by stopping all VFs at the same time */
- ice_for_each_vf(pf, i)
- ice_dis_vf_qs(&pf->vf[i]);
-
tmp = pf->num_alloc_vfs;
pf->num_qps_per_vf = 0;
pf->num_alloc_vfs = 0;
for (i = 0; i < tmp; i++) {
- if (test_bit(ICE_VF_STATE_INIT, pf->vf[i].vf_states)) {
+ struct ice_vf *vf = &pf->vf[i];
+
+ mutex_lock(&vf->cfg_lock);
+
+ ice_dis_vf_qs(vf);
+
+ if (test_bit(ICE_VF_STATE_INIT, vf->vf_states)) {
/* 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]);
+ ice_dis_vf_mappings(vf);
+ set_bit(ICE_VF_STATE_DIS, vf->vf_states);
+ ice_free_vf_res(vf);
}

- mutex_destroy(&pf->vf[i].cfg_lock);
+ mutex_unlock(&vf->cfg_lock);
+
+ mutex_destroy(&vf->cfg_lock);
}

if (ice_sriov_free_msix_res(pf))
@@ -1223,9 +1227,13 @@ bool ice_reset_all_vfs(struct ice_pf *pf
ice_for_each_vf(pf, v) {
vf = &pf->vf[v];

+ mutex_lock(&vf->cfg_lock);
+
ice_vf_pre_vsi_rebuild(vf);
ice_vf_rebuild_vsi(vf);
ice_vf_post_vsi_rebuild(vf);
+
+ mutex_unlock(&vf->cfg_lock);
}

ice_flush(hw);
@@ -1272,6 +1280,8 @@ bool ice_reset_vf(struct ice_vf *vf, boo
u32 reg;
int i;

+ lockdep_assert_held(&vf->cfg_lock);
+
dev = ice_pf_to_dev(pf);

if (test_bit(__ICE_VF_RESETS_DISABLED, pf->state)) {
@@ -1725,9 +1735,12 @@ void ice_process_vflr_event(struct ice_p
bit_idx = (hw->func_caps.vf_base_id + vf_id) % 32;
/* read GLGEN_VFLRSTAT register to find out the flr VFs */
reg = rd32(hw, GLGEN_VFLRSTAT(reg_idx));
- if (reg & BIT(bit_idx))
+ if (reg & BIT(bit_idx)) {
/* GLGEN_VFLRSTAT bit will be cleared in ice_reset_vf */
+ mutex_lock(&vf->cfg_lock);
ice_reset_vf(vf, true);
+ mutex_unlock(&vf->cfg_lock);
+ }
}
}

@@ -1804,7 +1817,9 @@ ice_vf_lan_overflow_event(struct ice_pf
if (!vf)
return;

+ mutex_lock(&vf->cfg_lock);
ice_vc_reset_vf(vf);
+ mutex_unlock(&vf->cfg_lock);
}

/**

\
 
 \ /
  Last update: 2022-03-07 10:56    [W:0.291 / U:0.208 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site