From 90e97538e6ab6359f8ea8f4ffe741dda4f536d42 Mon Sep 17 00:00:00 2001 From: Kalesh AP Date: Thu, 17 Apr 2025 10:24:47 -0700 Subject: bnxt_en: Remove unused field "ref_count" in struct bnxt_ulp [ Upstream commit 5bccacb4cc32cb835fe2fe100a210332c494e81d ] The "ref_count" field in struct bnxt_ulp is unused after commit a43c26fa2e6c ("RDMA/bnxt_re: Remove the sriov config callback"). So we can just remove it now. Reviewed-by: Somnath Kotur Signed-off-by: Kalesh AP Signed-off-by: Michael Chan Link: https://patch.msgid.link/20250417172448.1206107-4-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin --- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 5 ----- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 1 - 2 files changed, 6 deletions(-) (limited to 'drivers/net/ethernet/broadcom') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index 7564705d6478..84c4812414fd 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -149,7 +149,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) struct net_device *dev = edev->net; struct bnxt *bp = netdev_priv(dev); struct bnxt_ulp *ulp; - int i = 0; ulp = edev->ulp_tbl; netdev_lock(dev); @@ -165,10 +164,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev) synchronize_rcu(); ulp->max_async_event_id = 0; ulp->async_events_bmap = NULL; - while (atomic_read(&ulp->ref_count) != 0 && i < 10) { - msleep(100); - i++; - } mutex_unlock(&edev->en_dev_lock); netdev_unlock(dev); return; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h index 7fa3b8d1ebd2..f6b5efb5e775 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h @@ -50,7 +50,6 @@ struct bnxt_ulp { unsigned long *async_events_bmap; u16 max_async_event_id; u16 msix_requested; - atomic_t ref_count; }; struct bnxt_en_dev { -- cgit v1.2.3 From cca4ffd4a5b768752b91f6f75d77928e47886913 Mon Sep 17 00:00:00 2001 From: Kalesh AP Date: Fri, 13 Jun 2025 16:18:39 -0700 Subject: bnxt_en: Fix double invocation of bnxt_ulp_stop()/bnxt_ulp_start() [ Upstream commit 1e9ac33fa271be0d2480fd732f9642d81542500b ] Before the commit under the Fixes tag below, bnxt_ulp_stop() and bnxt_ulp_start() were always invoked in pairs. After that commit, the new bnxt_ulp_restart() can be invoked after bnxt_ulp_stop() has been called. This may result in the RoCE driver's aux driver .suspend() method being invoked twice. The 2nd bnxt_re_suspend() call will crash when it dereferences a NULL pointer: (NULL ib_device): Handle device suspend call BUG: kernel NULL pointer dereference, address: 0000000000000b78 PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 20 UID: 0 PID: 181 Comm: kworker/u96:5 Tainted: G S 6.15.0-rc1 #4 PREEMPT(voluntary) Tainted: [S]=CPU_OUT_OF_SPEC Hardware name: Dell Inc. PowerEdge R730/072T6D, BIOS 2.4.3 01/17/2017 Workqueue: bnxt_pf_wq bnxt_sp_task [bnxt_en] RIP: 0010:bnxt_re_suspend+0x45/0x1f0 [bnxt_re] Code: 8b 05 a7 3c 5b f5 48 89 44 24 18 31 c0 49 8b 5c 24 08 4d 8b 2c 24 e8 ea 06 0a f4 48 c7 c6 04 60 52 c0 48 89 df e8 1b ce f9 ff <48> 8b 83 78 0b 00 00 48 8b 80 38 03 00 00 a8 40 0f 85 b5 00 00 00 RSP: 0018:ffffa2e84084fd88 EFLAGS: 00010246 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000001 RDX: 0000000000000000 RSI: ffffffffb4b6b934 RDI: 00000000ffffffff RBP: ffffa1760954c9c0 R08: 0000000000000000 R09: c0000000ffffdfff R10: 0000000000000001 R11: ffffa2e84084fb50 R12: ffffa176031ef070 R13: ffffa17609775000 R14: ffffa17603adc180 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffffa17daa397000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000b78 CR3: 00000004aaa30003 CR4: 00000000003706f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: bnxt_ulp_stop+0x69/0x90 [bnxt_en] bnxt_sp_task+0x678/0x920 [bnxt_en] ? __schedule+0x514/0xf50 process_scheduled_works+0x9d/0x400 worker_thread+0x11c/0x260 ? __pfx_worker_thread+0x10/0x10 kthread+0xfe/0x1e0 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x2b/0x40 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 Check the BNXT_EN_FLAG_ULP_STOPPED flag and do not proceed if the flag is already set. This will preserve the original symmetrical bnxt_ulp_stop() and bnxt_ulp_start(). Also, inside bnxt_ulp_start(), clear the BNXT_EN_FLAG_ULP_STOPPED flag after taking the mutex to avoid any race condition. And for symmetry, only proceed in bnxt_ulp_start() if the BNXT_EN_FLAG_ULP_STOPPED is set. Fixes: 3c163f35bd50 ("bnxt_en: Optimize recovery path ULP locking in the driver") Signed-off-by: Kalesh AP Co-developed-by: Michael Chan Signed-off-by: Michael Chan Reviewed-by: Simon Horman Link: https://patch.msgid.link/20250613231841.377988-2-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin --- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) (limited to 'drivers/net/ethernet/broadcom') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index 84c4812414fd..2450a369b792 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -231,10 +231,9 @@ void bnxt_ulp_stop(struct bnxt *bp) return; mutex_lock(&edev->en_dev_lock); - if (!bnxt_ulp_registered(edev)) { - mutex_unlock(&edev->en_dev_lock); - return; - } + if (!bnxt_ulp_registered(edev) || + (edev->flags & BNXT_EN_FLAG_ULP_STOPPED)) + goto ulp_stop_exit; edev->flags |= BNXT_EN_FLAG_ULP_STOPPED; if (aux_priv) { @@ -250,6 +249,7 @@ void bnxt_ulp_stop(struct bnxt *bp) adrv->suspend(adev, pm); } } +ulp_stop_exit: mutex_unlock(&edev->en_dev_lock); } @@ -258,19 +258,13 @@ void bnxt_ulp_start(struct bnxt *bp, int err) struct bnxt_aux_priv *aux_priv = bp->aux_priv; struct bnxt_en_dev *edev = bp->edev; - if (!edev) - return; - - edev->flags &= ~BNXT_EN_FLAG_ULP_STOPPED; - - if (err) + if (!edev || err) return; mutex_lock(&edev->en_dev_lock); - if (!bnxt_ulp_registered(edev)) { - mutex_unlock(&edev->en_dev_lock); - return; - } + if (!bnxt_ulp_registered(edev) || + !(edev->flags & BNXT_EN_FLAG_ULP_STOPPED)) + goto ulp_start_exit; if (edev->ulp_tbl->msix_requested) bnxt_fill_msix_vecs(bp, edev->msix_entries); @@ -287,6 +281,8 @@ void bnxt_ulp_start(struct bnxt *bp, int err) adrv->resume(adev); } } +ulp_start_exit: + edev->flags &= ~BNXT_EN_FLAG_ULP_STOPPED; mutex_unlock(&edev->en_dev_lock); } -- cgit v1.2.3 From 6b69658f49c7d80e0fd2d2177f77201cd272aee1 Mon Sep 17 00:00:00 2001 From: Pavan Chebbi Date: Fri, 13 Jun 2025 16:18:40 -0700 Subject: bnxt_en: Add a helper function to configure MRU and RSS [ Upstream commit e11baaea94e2923739a98abeee85eb0667c04fd3 ] Add a new helper function that will configure MRU and RSS table of a VNIC. This will be useful when we configure both on a VNIC when resetting an RX ring. This function will be used again in the next bug fix patch where we have to reconfigure VNICs for RSS contexts. Suggested-by: Michael Chan Reviewed-by: Simon Horman Reviewed-by: David Wei Signed-off-by: Pavan Chebbi Signed-off-by: Michael Chan Link: https://patch.msgid.link/20250613231841.377988-3-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski Stable-dep-of: 5dacc94c6fe6 ("bnxt_en: Update MRU and RSS table of RSS contexts on queue reset") Signed-off-by: Sasha Levin --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 37 ++++++++++++++++++++++--------- 1 file changed, 26 insertions(+), 11 deletions(-) (limited to 'drivers/net/ethernet/broadcom') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 6afc2ab6fad2..3d975e50f943 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -10738,6 +10738,26 @@ void bnxt_del_one_rss_ctx(struct bnxt *bp, struct bnxt_rss_ctx *rss_ctx, bp->num_rss_ctx--; } +static int bnxt_set_vnic_mru_p5(struct bnxt *bp, struct bnxt_vnic_info *vnic, + u16 mru) +{ + int rc; + + if (mru) { + rc = bnxt_hwrm_vnic_set_rss_p5(bp, vnic, true); + if (rc) { + netdev_err(bp->dev, "hwrm vnic %d set rss failure rc: %d\n", + vnic->vnic_id, rc); + return rc; + } + } + vnic->mru = mru; + bnxt_hwrm_vnic_update(bp, vnic, + VNIC_UPDATE_REQ_ENABLES_MRU_VALID); + + return 0; +} + static void bnxt_hwrm_realloc_rss_ctx_vnic(struct bnxt *bp) { bool set_tpa = !!(bp->flags & BNXT_FLAG_TPA); @@ -15884,6 +15904,7 @@ static int bnxt_queue_start(struct net_device *dev, void *qmem, int idx) struct bnxt_vnic_info *vnic; struct bnxt_napi *bnapi; int i, rc; + u16 mru; rxr = &bp->rx_ring[idx]; clone = qmem; @@ -15933,18 +15954,13 @@ static int bnxt_queue_start(struct net_device *dev, void *qmem, int idx) napi_enable_locked(&bnapi->napi); bnxt_db_nq_arm(bp, &cpr->cp_db, cpr->cp_raw_cons); + mru = bp->dev->mtu + ETH_HLEN + VLAN_HLEN; for (i = 0; i < bp->nr_vnics; i++) { vnic = &bp->vnic_info[i]; - rc = bnxt_hwrm_vnic_set_rss_p5(bp, vnic, true); - if (rc) { - netdev_err(bp->dev, "hwrm vnic %d set rss failure rc: %d\n", - vnic->vnic_id, rc); + rc = bnxt_set_vnic_mru_p5(bp, vnic, mru); + if (rc) return rc; - } - vnic->mru = bp->dev->mtu + ETH_HLEN + VLAN_HLEN; - bnxt_hwrm_vnic_update(bp, vnic, - VNIC_UPDATE_REQ_ENABLES_MRU_VALID); } return 0; @@ -15969,9 +15985,8 @@ static int bnxt_queue_stop(struct net_device *dev, void *qmem, int idx) for (i = 0; i < bp->nr_vnics; i++) { vnic = &bp->vnic_info[i]; - vnic->mru = 0; - bnxt_hwrm_vnic_update(bp, vnic, - VNIC_UPDATE_REQ_ENABLES_MRU_VALID); + + bnxt_set_vnic_mru_p5(bp, vnic, 0); } /* Make sure NAPI sees that the VNIC is disabled */ synchronize_net(); -- cgit v1.2.3 From d44882e0c50574057f3cb0c7ef7b6e7568b85988 Mon Sep 17 00:00:00 2001 From: Pavan Chebbi Date: Fri, 13 Jun 2025 16:18:41 -0700 Subject: bnxt_en: Update MRU and RSS table of RSS contexts on queue reset [ Upstream commit 5dacc94c6fe61cde6f700e95cf35af9944b022c4 ] The commit under the Fixes tag below which updates the VNICs' RSS and MRU during .ndo_queue_start(), needs to be extended to cover any non-default RSS contexts which have their own VNICs. Without this step, packets that are destined to a non-default RSS context may be dropped after .ndo_queue_start(). We further optimize this scheme by updating the VNIC only if the RX ring being restarted is in the RSS table of the VNIC. Updating the VNIC (in particular setting the MRU to 0) will momentarily stop all traffic to all rings in the RSS table. Any VNIC that has the RX ring excluded from the RSS table can skip this step and avoid the traffic disruption. Note that this scheme is just an improvement. A VNIC with multiple rings in the RSS table will still see traffic disruptions to all rings in the RSS table when one of the rings is being restarted. We are working on a FW scheme that will improve upon this further. Fixes: 5ac066b7b062 ("bnxt_en: Fix queue start to update vnic RSS table") Reported-by: David Wei Signed-off-by: Pavan Chebbi Signed-off-by: Michael Chan Link: https://patch.msgid.link/20250613231841.377988-4-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 56 ++++++++++++++++++++++++++++--- 1 file changed, 51 insertions(+), 5 deletions(-) (limited to 'drivers/net/ethernet/broadcom') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 3d975e50f943..c365a9e64f72 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -10738,11 +10738,39 @@ void bnxt_del_one_rss_ctx(struct bnxt *bp, struct bnxt_rss_ctx *rss_ctx, bp->num_rss_ctx--; } +static bool bnxt_vnic_has_rx_ring(struct bnxt *bp, struct bnxt_vnic_info *vnic, + int rxr_id) +{ + u16 tbl_size = bnxt_get_rxfh_indir_size(bp->dev); + int i, vnic_rx; + + /* Ntuple VNIC always has all the rx rings. Any change of ring id + * must be updated because a future filter may use it. + */ + if (vnic->flags & BNXT_VNIC_NTUPLE_FLAG) + return true; + + for (i = 0; i < tbl_size; i++) { + if (vnic->flags & BNXT_VNIC_RSSCTX_FLAG) + vnic_rx = ethtool_rxfh_context_indir(vnic->rss_ctx)[i]; + else + vnic_rx = bp->rss_indir_tbl[i]; + + if (rxr_id == vnic_rx) + return true; + } + + return false; +} + static int bnxt_set_vnic_mru_p5(struct bnxt *bp, struct bnxt_vnic_info *vnic, - u16 mru) + u16 mru, int rxr_id) { int rc; + if (!bnxt_vnic_has_rx_ring(bp, vnic, rxr_id)) + return 0; + if (mru) { rc = bnxt_hwrm_vnic_set_rss_p5(bp, vnic, true); if (rc) { @@ -10758,6 +10786,24 @@ static int bnxt_set_vnic_mru_p5(struct bnxt *bp, struct bnxt_vnic_info *vnic, return 0; } +static int bnxt_set_rss_ctx_vnic_mru(struct bnxt *bp, u16 mru, int rxr_id) +{ + struct ethtool_rxfh_context *ctx; + unsigned long context; + int rc; + + xa_for_each(&bp->dev->ethtool->rss_ctx, context, ctx) { + struct bnxt_rss_ctx *rss_ctx = ethtool_rxfh_context_priv(ctx); + struct bnxt_vnic_info *vnic = &rss_ctx->vnic; + + rc = bnxt_set_vnic_mru_p5(bp, vnic, mru, rxr_id); + if (rc) + return rc; + } + + return 0; +} + static void bnxt_hwrm_realloc_rss_ctx_vnic(struct bnxt *bp) { bool set_tpa = !!(bp->flags & BNXT_FLAG_TPA); @@ -15958,12 +16004,11 @@ static int bnxt_queue_start(struct net_device *dev, void *qmem, int idx) for (i = 0; i < bp->nr_vnics; i++) { vnic = &bp->vnic_info[i]; - rc = bnxt_set_vnic_mru_p5(bp, vnic, mru); + rc = bnxt_set_vnic_mru_p5(bp, vnic, mru, idx); if (rc) return rc; } - - return 0; + return bnxt_set_rss_ctx_vnic_mru(bp, mru, idx); err_reset: netdev_err(bp->dev, "Unexpected HWRM error during queue start rc: %d\n", @@ -15986,8 +16031,9 @@ static int bnxt_queue_stop(struct net_device *dev, void *qmem, int idx) for (i = 0; i < bp->nr_vnics; i++) { vnic = &bp->vnic_info[i]; - bnxt_set_vnic_mru_p5(bp, vnic, 0); + bnxt_set_vnic_mru_p5(bp, vnic, 0, idx); } + bnxt_set_rss_ctx_vnic_mru(bp, 0, idx); /* Make sure NAPI sees that the VNIC is disabled */ synchronize_net(); rxr = &bp->rx_ring[idx]; -- cgit v1.2.3