May 2025 - Linux-stable-mirror

RE: [PATCH v6] virtio_blk: Fix disk deletion hang on device surprise removal

by Parav Pandit

> From: Parav Pandit <parav(a)nvidia.com> > Sent: Thursday, May 22, 2025 1:19 PM > To: Max Gurtovoy <mgurtovoy(a)nvidia.com>; Israel Rukshin > <israelr(a)nvidia.com> > Cc: Parav Pandit <parav(a)nvidia.com>; stable(a)vger.kernel.org; NBU-Contact- > Li Rongqing (EXTERNAL) <lirongqing(a)baidu.com> > Subject: [PATCH v6] virtio_blk: Fix disk deletion hang on device surprise > removal > > When the PCI device is surprise removed, requests may not complete the > device as the VQ is marked as broken. Due to this, the disk deletion hangs. > > Fix it by aborting the requests when the VQ is broken. > > With this fix now fio completes swiftly. > An alternative of IO timeout has been considered, however when the driver > knows about unresponsive block device, swiftly clearing them enables users > and upper layers to react quickly. > > Verified with multiple device unplug iterations with pending requests in virtio > used ring and some pending with the device. > > Fixes: 43bb40c5b926 ("virtio_pci: Support surprise removal of virtio pci > device") > Cc: stable(a)vger.kernel.org > Reported-by: Li RongQing <lirongqing(a)baidu.com> > Closes: > https://lore.kernel.org/virtualization/c45dd68698cd47238c55fb73ca9b4741 > @baidu.com/ > Signed-off-by: Parav Pandit <parav(a)nvidia.com> > This is an internal patch, which got CCed to stable by mistake. Please ignore this patch for stable kernels. It is still under internal review. I am sorry for the noise. > --- > v1->v2: (internal v5->v6): > - Addressed comments from Stephan > - fixed spelling to 'waiting' > v1->v2: (internal v4->v5): > - Addressed comments from MST > - removed the vq broken check in queue_rq(s) > --- > drivers/block/virtio_blk.c | 85 > ++++++++++++++++++++++++++++++++++++++ > 1 file changed, 85 insertions(+) > > diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index > 7cffea01d868..04f24ec20405 100644 > --- a/drivers/block/virtio_blk.c > +++ b/drivers/block/virtio_blk.c > @@ -1554,6 +1554,89 @@ static int virtblk_probe(struct virtio_device *vdev) > return err; > } > > +static bool virtblk_request_cancel(struct request *rq, void *data) { > + struct virtblk_req *vbr = blk_mq_rq_to_pdu(rq); > + struct virtio_blk *vblk = data; > + struct virtio_blk_vq *vq; > + unsigned long flags; > + > + vq = &vblk->vqs[rq->mq_hctx->queue_num]; > + > + spin_lock_irqsave(&vq->lock, flags); > + > + vbr->in_hdr.status = VIRTIO_BLK_S_IOERR; > + if (blk_mq_request_started(rq) && !blk_mq_request_completed(rq)) > + blk_mq_complete_request(rq); > + > + spin_unlock_irqrestore(&vq->lock, flags); > + return true; > +} > + > +static void virtblk_broken_device_cleanup(struct virtio_blk *vblk) { > + struct request_queue *q = vblk->disk->queue; > + > + return; > + > + if (!virtqueue_is_broken(vblk->vqs[0].vq)) > + return; > + > + /* Start freezing the queue, so that new requests keeps waiting at the > + * door of bio_queue_enter(). We cannot fully freeze the queue > because > + * freezed queue is an empty queue and there are pending requests, > so > + * only start freezing it. > + */ > + blk_freeze_queue_start(q); > + > + /* When quiescing completes, all ongoing dispatches have completed > + * and no new dispatch will happen towards the driver. > + * This ensures that later when cancel is attempted, then are not > + * getting processed by the queue_rq() or queue_rqs() handlers. > + */ > + blk_mq_quiesce_queue(q); > + > + /* > + * Synchronize with any ongoing VQ callbacks, effectively quiescing > + * the device and preventing it from completing further requests > + * to the block layer. Any outstanding, incomplete requests will be > + * completed by virtblk_request_cancel(). > + */ > + virtio_synchronize_cbs(vblk->vdev); > + > + /* At this point, no new requests can enter the queue_rq() and > + * completion routine will not complete any new requests either for > the > + * broken vq. Hence, it is safe to cancel all requests which are > + * started. > + */ > + blk_mq_tagset_busy_iter(&vblk->tag_set, virtblk_request_cancel, > vblk); > + blk_mq_tagset_wait_completed_request(&vblk->tag_set); > + > + /* All pending requests are cleaned up. Time to resume so that disk > + * deletion can be smooth. Start the HW queues so that when queue is > + * unquiesced requests can again enter the driver. > + */ > + blk_mq_start_stopped_hw_queues(q, true); > + > + /* Unquiescing will trigger dispatching any pending requests to the > + * driver which has crossed bio_queue_enter() to the driver. > + */ > + blk_mq_unquiesce_queue(q); > + > + /* Wait for all pending dispatches to terminate which may have been > + * initiated after unquiescing. > + */ > + blk_mq_freeze_queue_wait(q); > + > + /* Mark the disk dead so that once queue unfreeze, the requests > + * waiting at the door of bio_queue_enter() can be aborted right away. > + */ > + blk_mark_disk_dead(vblk->disk); > + > + /* Unfreeze the queue so that any waiting requests will be aborted. */ > + blk_mq_unfreeze_queue_nomemrestore(q); > +} > + > static void virtblk_remove(struct virtio_device *vdev) { > struct virtio_blk *vblk = vdev->priv; > @@ -1561,6 +1644,8 @@ static void virtblk_remove(struct virtio_device > *vdev) > /* Make sure no work handler is accessing the device. */ > flush_work(&vblk->config_work); > > + virtblk_broken_device_cleanup(vblk); > + > del_gendisk(vblk->disk); > blk_mq_free_tag_set(&vblk->tag_set); > > -- > 2.34.1

1 month, 2 weeks

1
0
0 0

[PATCH net 4/4] can: slcan: allow reception of short error messages

by Marc Kleine-Budde

From: Carlos Sanchez <carlossanchez(a)geotab.com> Allows slcan to receive short messages (typically errors) from the serial interface. When error support was added to slcan protocol in b32ff4668544e1333b694fcc7812b2d7397b4d6a ("can: slcan: extend the protocol with error info") the minimum valid message size changed from 5 (minimum standard can frame tIII0) to 3 ("e1a" is a valid protocol message, it is one of the examples given in the comments for slcan_bump_err() ), but the check for minimum message length prodicating all decoding was not adjusted. This makes short error messages discarded and error frames not being generated. This patch changes the minimum length to the new minimum (3 characters, excluding terminator, is now a valid message). Signed-off-by: Carlos Sanchez <carlossanchez(a)geotab.com> Fixes: b32ff4668544 ("can: slcan: extend the protocol with error info") Reviewed-by: Vincent Mailhol <mailhol.vincent(a)wanadoo.fr> Link: https://patch.msgid.link/20250520102305.1097494-1-carlossanchez@geotab.com Cc: stable(a)vger.kernel.org Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> --- drivers/net/can/slcan/slcan-core.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/drivers/net/can/slcan/slcan-core.c b/drivers/net/can/slcan/slcan-core.c index 24c6622d36bd..58ff2ec1d975 100644 --- a/drivers/net/can/slcan/slcan-core.c +++ b/drivers/net/can/slcan/slcan-core.c @@ -71,12 +71,21 @@ MODULE_AUTHOR("Dario Binacchi <dario.binacchi(a)amarulasolutions.com>"); #define SLCAN_CMD_LEN 1 #define SLCAN_SFF_ID_LEN 3 #define SLCAN_EFF_ID_LEN 8 +#define SLCAN_DATA_LENGTH_LEN 1 +#define SLCAN_ERROR_LEN 1 #define SLCAN_STATE_LEN 1 #define SLCAN_STATE_BE_RXCNT_LEN 3 #define SLCAN_STATE_BE_TXCNT_LEN 3 -#define SLCAN_STATE_FRAME_LEN (1 + SLCAN_CMD_LEN + \ - SLCAN_STATE_BE_RXCNT_LEN + \ - SLCAN_STATE_BE_TXCNT_LEN) +#define SLCAN_STATE_MSG_LEN (SLCAN_CMD_LEN + \ + SLCAN_STATE_LEN + \ + SLCAN_STATE_BE_RXCNT_LEN + \ + SLCAN_STATE_BE_TXCNT_LEN) +#define SLCAN_ERROR_MSG_LEN_MIN (SLCAN_CMD_LEN + \ + SLCAN_ERROR_LEN + \ + SLCAN_DATA_LENGTH_LEN) +#define SLCAN_FRAME_MSG_LEN_MIN (SLCAN_CMD_LEN + \ + SLCAN_SFF_ID_LEN + \ + SLCAN_DATA_LENGTH_LEN) struct slcan { struct can_priv can; @@ -176,6 +185,9 @@ static void slcan_bump_frame(struct slcan *sl) u32 tmpid; char *cmd = sl->rbuff; + if (sl->rcount < SLCAN_FRAME_MSG_LEN_MIN) + return; + skb = alloc_can_skb(sl->dev, &cf); if (unlikely(!skb)) { sl->dev->stats.rx_dropped++; @@ -281,7 +293,7 @@ static void slcan_bump_state(struct slcan *sl) return; } - if (state == sl->can.state || sl->rcount < SLCAN_STATE_FRAME_LEN) + if (state == sl->can.state || sl->rcount != SLCAN_STATE_MSG_LEN) return; cmd += SLCAN_STATE_BE_RXCNT_LEN + SLCAN_CMD_LEN + 1; @@ -328,6 +340,9 @@ static void slcan_bump_err(struct slcan *sl) bool rx_errors = false, tx_errors = false, rx_over_errors = false; int i, len; + if (sl->rcount < SLCAN_ERROR_MSG_LEN_MIN) + return; + /* get len from sanitized ASCII value */ len = cmd[1]; if (len >= '0' && len < '9') @@ -456,8 +471,7 @@ static void slcan_bump(struct slcan *sl) static void slcan_unesc(struct slcan *sl, unsigned char s) { if ((s == '\r') || (s == '\a')) { /* CR or BEL ends the pdu */ - if (!test_and_clear_bit(SLF_ERROR, &sl->flags) && - sl->rcount > 4) + if (!test_and_clear_bit(SLF_ERROR, &sl->flags)) slcan_bump(sl); sl->rcount = 0; -- 2.47.2

1 month, 2 weeks

1
0
0 0

[PATCH net 3/4] can: kvaser_pciefd: Continue parsing DMA buf after dropped RX

by Marc Kleine-Budde

From: Axel Forsman <axfo(a)kvaser.com> Going bus-off on a channel doing RX could result in dropped packets. As netif_running() gets cleared before the channel abort procedure, the handling of any last RDATA packets would see netif_rx() return non-zero to signal a dropped packet. kvaser_pciefd_read_buffer() dealt with this "error" by breaking out of processing the remaining DMA RX buffer. Only return an error from kvaser_pciefd_read_buffer() due to packet corruption, otherwise handle it internally. Cc: stable(a)vger.kernel.org Signed-off-by: Axel Forsman <axfo(a)kvaser.com> Tested-by: Jimmy Assarsson <extja(a)kvaser.com> Reviewed-by: Jimmy Assarsson <extja(a)kvaser.com> Link: https://patch.msgid.link/20250520114332.8961-4-axfo@kvaser.com Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> --- drivers/net/can/kvaser_pciefd.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/net/can/kvaser_pciefd.c b/drivers/net/can/kvaser_pciefd.c index a61cbade96d9..f6921368cd14 100644 --- a/drivers/net/can/kvaser_pciefd.c +++ b/drivers/net/can/kvaser_pciefd.c @@ -1209,7 +1209,7 @@ static int kvaser_pciefd_handle_data_packet(struct kvaser_pciefd *pcie, skb = alloc_canfd_skb(priv->dev, &cf); if (!skb) { priv->dev->stats.rx_dropped++; - return -ENOMEM; + return 0; } cf->len = can_fd_dlc2len(dlc); @@ -1221,7 +1221,7 @@ static int kvaser_pciefd_handle_data_packet(struct kvaser_pciefd *pcie, skb = alloc_can_skb(priv->dev, (struct can_frame **)&cf); if (!skb) { priv->dev->stats.rx_dropped++; - return -ENOMEM; + return 0; } can_frame_set_cc_len((struct can_frame *)cf, dlc, priv->ctrlmode); } @@ -1239,7 +1239,9 @@ static int kvaser_pciefd_handle_data_packet(struct kvaser_pciefd *pcie, priv->dev->stats.rx_packets++; kvaser_pciefd_set_skb_timestamp(pcie, skb, p->timestamp); - return netif_rx(skb); + netif_rx(skb); + + return 0; } static void kvaser_pciefd_change_state(struct kvaser_pciefd_can *can, -- 2.47.2

1 month, 2 weeks

1
0
0 0

[PATCH net 2/4] can: kvaser_pciefd: Fix echo_skb race

by Marc Kleine-Budde

From: Axel Forsman <axfo(a)kvaser.com> The functions kvaser_pciefd_start_xmit() and kvaser_pciefd_handle_ack_packet() raced to stop/wake TX queues and get/put echo skbs, as kvaser_pciefd_can->echo_lock was only ever taken when transmitting and KCAN_TX_NR_PACKETS_CURRENT gets decremented prior to handling of ACKs. E.g., this caused the following error: can_put_echo_skb: BUG! echo_skb 5 is occupied! Instead, use the synchronization helpers in netdev_queues.h. As those piggyback on BQL barriers, start updating in-flight packets and bytes counts as well. Cc: stable(a)vger.kernel.org Signed-off-by: Axel Forsman <axfo(a)kvaser.com> Tested-by: Jimmy Assarsson <extja(a)kvaser.com> Reviewed-by: Jimmy Assarsson <extja(a)kvaser.com> Link: https://patch.msgid.link/20250520114332.8961-3-axfo@kvaser.com Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> --- drivers/net/can/kvaser_pciefd.c | 93 +++++++++++++++++++++------------ 1 file changed, 59 insertions(+), 34 deletions(-) diff --git a/drivers/net/can/kvaser_pciefd.c b/drivers/net/can/kvaser_pciefd.c index 9cc9176c2058..a61cbade96d9 100644 --- a/drivers/net/can/kvaser_pciefd.c +++ b/drivers/net/can/kvaser_pciefd.c @@ -16,6 +16,7 @@ #include <linux/netdevice.h> #include <linux/pci.h> #include <linux/timer.h> +#include <net/netdev_queues.h> MODULE_LICENSE("Dual BSD/GPL"); MODULE_AUTHOR("Kvaser AB <support(a)kvaser.com>"); @@ -410,10 +411,13 @@ struct kvaser_pciefd_can { void __iomem *reg_base; struct can_berr_counter bec; u8 cmd_seq; + u8 tx_max_count; + u8 tx_idx; + u8 ack_idx; int err_rep_cnt; - int echo_idx; + unsigned int completed_tx_pkts; + unsigned int completed_tx_bytes; spinlock_t lock; /* Locks sensitive registers (e.g. MODE) */ - spinlock_t echo_lock; /* Locks the message echo buffer */ struct timer_list bec_poll_timer; struct completion start_comp, flush_comp; }; @@ -714,6 +718,9 @@ static int kvaser_pciefd_open(struct net_device *netdev) int ret; struct kvaser_pciefd_can *can = netdev_priv(netdev); + can->tx_idx = 0; + can->ack_idx = 0; + ret = open_candev(netdev); if (ret) return ret; @@ -745,21 +752,26 @@ static int kvaser_pciefd_stop(struct net_device *netdev) timer_delete(&can->bec_poll_timer); } can->can.state = CAN_STATE_STOPPED; + netdev_reset_queue(netdev); close_candev(netdev); return ret; } +static unsigned int kvaser_pciefd_tx_avail(const struct kvaser_pciefd_can *can) +{ + return can->tx_max_count - (READ_ONCE(can->tx_idx) - READ_ONCE(can->ack_idx)); +} + static int kvaser_pciefd_prepare_tx_packet(struct kvaser_pciefd_tx_packet *p, - struct kvaser_pciefd_can *can, + struct can_priv *can, u8 seq, struct sk_buff *skb) { struct canfd_frame *cf = (struct canfd_frame *)skb->data; int packet_size; - int seq = can->echo_idx; memset(p, 0, sizeof(*p)); - if (can->can.ctrlmode & CAN_CTRLMODE_ONE_SHOT) + if (can->ctrlmode & CAN_CTRLMODE_ONE_SHOT) p->header[1] |= KVASER_PCIEFD_TPACKET_SMS; if (cf->can_id & CAN_RTR_FLAG) @@ -782,7 +794,7 @@ static int kvaser_pciefd_prepare_tx_packet(struct kvaser_pciefd_tx_packet *p, } else { p->header[1] |= FIELD_PREP(KVASER_PCIEFD_RPACKET_DLC_MASK, - can_get_cc_dlc((struct can_frame *)cf, can->can.ctrlmode)); + can_get_cc_dlc((struct can_frame *)cf, can->ctrlmode)); } p->header[1] |= FIELD_PREP(KVASER_PCIEFD_PACKET_SEQ_MASK, seq); @@ -797,22 +809,24 @@ static netdev_tx_t kvaser_pciefd_start_xmit(struct sk_buff *skb, struct net_device *netdev) { struct kvaser_pciefd_can *can = netdev_priv(netdev); - unsigned long irq_flags; struct kvaser_pciefd_tx_packet packet; + unsigned int seq = can->tx_idx & (can->can.echo_skb_max - 1); + unsigned int frame_len; int nr_words; - u8 count; if (can_dev_dropped_skb(netdev, skb)) return NETDEV_TX_OK; + if (!netif_subqueue_maybe_stop(netdev, 0, kvaser_pciefd_tx_avail(can), 1, 1)) + return NETDEV_TX_BUSY; - nr_words = kvaser_pciefd_prepare_tx_packet(&packet, can, skb); + nr_words = kvaser_pciefd_prepare_tx_packet(&packet, &can->can, seq, skb); - spin_lock_irqsave(&can->echo_lock, irq_flags); /* Prepare and save echo skb in internal slot */ - can_put_echo_skb(skb, netdev, can->echo_idx, 0); - - /* Move echo index to the next slot */ - can->echo_idx = (can->echo_idx + 1) % can->can.echo_skb_max; + WRITE_ONCE(can->can.echo_skb[seq], NULL); + frame_len = can_skb_get_frame_len(skb); + can_put_echo_skb(skb, netdev, seq, frame_len); + netdev_sent_queue(netdev, frame_len); + WRITE_ONCE(can->tx_idx, can->tx_idx + 1); /* Write header to fifo */ iowrite32(packet.header[0], @@ -836,14 +850,7 @@ static netdev_tx_t kvaser_pciefd_start_xmit(struct sk_buff *skb, KVASER_PCIEFD_KCAN_FIFO_LAST_REG); } - count = FIELD_GET(KVASER_PCIEFD_KCAN_TX_NR_PACKETS_CURRENT_MASK, - ioread32(can->reg_base + KVASER_PCIEFD_KCAN_TX_NR_PACKETS_REG)); - /* No room for a new message, stop the queue until at least one - * successful transmit - */ - if (count >= can->can.echo_skb_max || can->can.echo_skb[can->echo_idx]) - netif_stop_queue(netdev); - spin_unlock_irqrestore(&can->echo_lock, irq_flags); + netif_subqueue_maybe_stop(netdev, 0, kvaser_pciefd_tx_avail(can), 1, 1); return NETDEV_TX_OK; } @@ -970,6 +977,8 @@ static int kvaser_pciefd_setup_can_ctrls(struct kvaser_pciefd *pcie) can->kv_pcie = pcie; can->cmd_seq = 0; can->err_rep_cnt = 0; + can->completed_tx_pkts = 0; + can->completed_tx_bytes = 0; can->bec.txerr = 0; can->bec.rxerr = 0; @@ -983,11 +992,10 @@ static int kvaser_pciefd_setup_can_ctrls(struct kvaser_pciefd *pcie) tx_nr_packets_max = FIELD_GET(KVASER_PCIEFD_KCAN_TX_NR_PACKETS_MAX_MASK, ioread32(can->reg_base + KVASER_PCIEFD_KCAN_TX_NR_PACKETS_REG)); + can->tx_max_count = min(KVASER_PCIEFD_CAN_TX_MAX_COUNT, tx_nr_packets_max - 1); can->can.clock.freq = pcie->freq; - can->can.echo_skb_max = min(KVASER_PCIEFD_CAN_TX_MAX_COUNT, tx_nr_packets_max - 1); - can->echo_idx = 0; - spin_lock_init(&can->echo_lock); + can->can.echo_skb_max = roundup_pow_of_two(can->tx_max_count); spin_lock_init(&can->lock); can->can.bittiming_const = &kvaser_pciefd_bittiming_const; @@ -1510,19 +1518,21 @@ static int kvaser_pciefd_handle_ack_packet(struct kvaser_pciefd *pcie, netdev_dbg(can->can.dev, "Packet was flushed\n"); } else { int echo_idx = FIELD_GET(KVASER_PCIEFD_PACKET_SEQ_MASK, p->header[0]); - int len; - u8 count; + unsigned int len, frame_len = 0; struct sk_buff *skb; + if (echo_idx != (can->ack_idx & (can->can.echo_skb_max - 1))) + return 0; skb = can->can.echo_skb[echo_idx]; - if (skb) - kvaser_pciefd_set_skb_timestamp(pcie, skb, p->timestamp); - len = can_get_echo_skb(can->can.dev, echo_idx, NULL); - count = FIELD_GET(KVASER_PCIEFD_KCAN_TX_NR_PACKETS_CURRENT_MASK, - ioread32(can->reg_base + KVASER_PCIEFD_KCAN_TX_NR_PACKETS_REG)); + if (!skb) + return 0; + kvaser_pciefd_set_skb_timestamp(pcie, skb, p->timestamp); + len = can_get_echo_skb(can->can.dev, echo_idx, &frame_len); - if (count < can->can.echo_skb_max && netif_queue_stopped(can->can.dev)) - netif_wake_queue(can->can.dev); + /* Pairs with barrier in kvaser_pciefd_start_xmit() */ + smp_store_release(&can->ack_idx, can->ack_idx + 1); + can->completed_tx_pkts++; + can->completed_tx_bytes += frame_len; if (!one_shot_fail) { can->can.dev->stats.tx_bytes += len; @@ -1638,11 +1648,26 @@ static int kvaser_pciefd_read_buffer(struct kvaser_pciefd *pcie, int dma_buf) { int pos = 0; int res = 0; + unsigned int i; do { res = kvaser_pciefd_read_packet(pcie, &pos, dma_buf); } while (!res && pos > 0 && pos < KVASER_PCIEFD_DMA_SIZE); + /* Report ACKs in this buffer to BQL en masse for correct periods */ + for (i = 0; i < pcie->nr_channels; ++i) { + struct kvaser_pciefd_can *can = pcie->can[i]; + + if (!can->completed_tx_pkts) + continue; + netif_subqueue_completed_wake(can->can.dev, 0, + can->completed_tx_pkts, + can->completed_tx_bytes, + kvaser_pciefd_tx_avail(can), 1); + can->completed_tx_pkts = 0; + can->completed_tx_bytes = 0; + } + return res; } -- 2.47.2

1 month, 2 weeks

1
0
0 0

[PATCH 5.10 1/1] ocfs2: fix deadlock in ocfs2_get_system_file_inode

by Dmitriy Privalov

From: Mohammed Anees <pvmohammedanees2003(a)gmail.com> commit 7bf1823e010e8db2fb649c790bd1b449a75f52d8 upstream. syzbot has found a possible deadlock in ocfs2_get_system_file_inode [1]. The scenario is depicted here, CPU0 CPU1 lock(&ocfs2_file_ip_alloc_sem_key); lock(&osb->system_file_mutex); lock(&ocfs2_file_ip_alloc_sem_key); lock(&osb->system_file_mutex); The function calls which could lead to this are: CPU0 ocfs2_mknod - lock(&ocfs2_file_ip_alloc_sem_key); . . . ocfs2_get_system_file_inode - lock(&osb->system_file_mutex); CPU1 - ocfs2_fill_super - lock(&osb->system_file_mutex); . . . ocfs2_read_virt_blocks - lock(&ocfs2_file_ip_alloc_sem_key); This issue can be resolved by making the down_read -> down_read_try in the ocfs2_read_virt_blocks. [1] https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd Link: https://lkml.kernel.org/r/20240924093257.7181-1-pvmohammedanees2003@gmail.c… Signed-off-by: Mohammed Anees <pvmohammedanees2003(a)gmail.com> Reviewed-by: Joseph Qi <joseph.qi(a)linux.alibaba.com> Reported-by: <syzbot+e0055ea09f1f5e6fabdd(a)syzkaller.appspotmail.com> Closes: https://syzkaller.appspot.com/bug?extid=e0055ea09f1f5e6fabdd Tested-by: syzbot+e0055ea09f1f5e6fabdd(a)syzkaller.appspotmail.com Cc: Mark Fasheh <mark(a)fasheh.com> Cc: Joel Becker <jlbec(a)evilplan.org> Cc: Junxiao Bi <junxiao.bi(a)oracle.com> Cc: Changwei Ge <gechangwei(a)live.cn> Cc: Gang He <ghe(a)suse.com> Cc: Jun Piao <piaojun(a)huawei.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Dmitriy Privalov <d.privalov(a)omp.ru> --- fs/ocfs2/extent_map.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c index 70a768b623cf..f7672472fa82 100644 --- a/fs/ocfs2/extent_map.c +++ b/fs/ocfs2/extent_map.c @@ -973,7 +973,13 @@ int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr, } while (done < nr) { - down_read(&OCFS2_I(inode)->ip_alloc_sem); + if (!down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem)) { + rc = -EAGAIN; + mlog(ML_ERROR, + "Inode #%llu ip_alloc_sem is temporarily unavailable\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); + break; + } rc = ocfs2_extent_map_get_blocks(inode, v_block + done, &p_block, &p_count, NULL); up_read(&OCFS2_I(inode)->ip_alloc_sem); -- 2.34.1

1 month, 2 weeks

1
0
0 0

[PATCH 1/2] drm/amdgpu/vcn4.0.5: split code along instances

by Eric Naim

From: Alex Deucher <alexander.deucher(a)amd.com> commit ecc9ab4e924b7eb9e2c4a668162aaa1d9d60d08c upstream Split the code on a per instance basis. This will allow us to use the per instance functions in the future to handle more things per instance. v2: squash in fix for stop() from Boyuan Reviewed-by: Boyuan Zhang <Boyuan.Zhang(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Tested-by: Eric Naim <dnaim(a)cachyos.org> Signed-off-by: Eric Naim <dnaim(a)cachyos.org> --- drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c | 444 ++++++++++++------------ 1 file changed, 220 insertions(+), 224 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c index d2dfdb141b24..9f9a9bf8dab9 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c @@ -991,183 +991,178 @@ static int vcn_v4_0_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, b * vcn_v4_0_5_start - VCN start * * @adev: amdgpu_device pointer + * @i: instance to start * * Start VCN block */ -static int vcn_v4_0_5_start(struct amdgpu_device *adev) +static int vcn_v4_0_5_start(struct amdgpu_device *adev, int i) { volatile struct amdgpu_vcn4_fw_shared *fw_shared; struct amdgpu_ring *ring; uint32_t tmp; - int i, j, k, r; + int j, k, r; - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { - if (adev->pm.dpm_enabled) - amdgpu_dpm_enable_vcn(adev, true, i); - } + if (adev->vcn.harvest_config & (1 << i)) + return 0; - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { - if (adev->vcn.harvest_config & (1 << i)) - continue; + if (adev->pm.dpm_enabled) + amdgpu_dpm_enable_vcn(adev, true, i); - fw_shared = adev->vcn.inst[i].fw_shared.cpu_addr; + fw_shared = adev->vcn.inst[i].fw_shared.cpu_addr; - if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) { - r = vcn_v4_0_5_start_dpg_mode(adev, i, adev->vcn.indirect_sram); - continue; - } + if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) + return vcn_v4_0_5_start_dpg_mode(adev, i, adev->vcn.indirect_sram); - /* disable VCN power gating */ - vcn_v4_0_5_disable_static_power_gating(adev, i); - - /* set VCN status busy */ - tmp = RREG32_SOC15(VCN, i, regUVD_STATUS) | UVD_STATUS__UVD_BUSY; - WREG32_SOC15(VCN, i, regUVD_STATUS, tmp); - - /*SW clock gating */ - vcn_v4_0_5_disable_clock_gating(adev, i); - - /* enable VCPU clock */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), - UVD_VCPU_CNTL__CLK_EN_MASK, ~UVD_VCPU_CNTL__CLK_EN_MASK); - - /* disable master interrupt */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_MASTINT_EN), 0, - ~UVD_MASTINT_EN__VCPU_EN_MASK); - - /* enable LMI MC and UMC channels */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_LMI_CTRL2), 0, - ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); - - tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); - tmp &= ~UVD_SOFT_RESET__LMI_SOFT_RESET_MASK; - tmp &= ~UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK; - WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); - - /* setup regUVD_LMI_CTRL */ - tmp = RREG32_SOC15(VCN, i, regUVD_LMI_CTRL); - WREG32_SOC15(VCN, i, regUVD_LMI_CTRL, tmp | - UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | - UVD_LMI_CTRL__MASK_MC_URGENT_MASK | - UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK | - UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK); - - /* setup regUVD_MPC_CNTL */ - tmp = RREG32_SOC15(VCN, i, regUVD_MPC_CNTL); - tmp &= ~UVD_MPC_CNTL__REPLACEMENT_MODE_MASK; - tmp |= 0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT; - WREG32_SOC15(VCN, i, regUVD_MPC_CNTL, tmp); - - /* setup UVD_MPC_SET_MUXA0 */ - WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUXA0, - ((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) | - (0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) | - (0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) | - (0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT))); - - /* setup UVD_MPC_SET_MUXB0 */ - WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUXB0, - ((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) | - (0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) | - (0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) | - (0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT))); - - /* setup UVD_MPC_SET_MUX */ - WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUX, - ((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) | - (0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) | - (0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT))); - - vcn_v4_0_5_mc_resume(adev, i); - - /* VCN global tiling registers */ - WREG32_SOC15(VCN, i, regUVD_GFX10_ADDR_CONFIG, - adev->gfx.config.gb_addr_config); - - /* unblock VCPU register access */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_RB_ARB_CTRL), 0, - ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); - - /* release VCPU reset to boot */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, - ~UVD_VCPU_CNTL__BLK_RST_MASK); - - for (j = 0; j < 10; ++j) { - uint32_t status; - - for (k = 0; k < 100; ++k) { - status = RREG32_SOC15(VCN, i, regUVD_STATUS); - if (status & 2) - break; - mdelay(10); - if (amdgpu_emu_mode == 1) - msleep(1); - } + /* disable VCN power gating */ + vcn_v4_0_5_disable_static_power_gating(adev, i); + + /* set VCN status busy */ + tmp = RREG32_SOC15(VCN, i, regUVD_STATUS) | UVD_STATUS__UVD_BUSY; + WREG32_SOC15(VCN, i, regUVD_STATUS, tmp); - if (amdgpu_emu_mode == 1) { - r = -1; - if (status & 2) { - r = 0; - break; - } - } else { + /* SW clock gating */ + vcn_v4_0_5_disable_clock_gating(adev, i); + + /* enable VCPU clock */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), + UVD_VCPU_CNTL__CLK_EN_MASK, ~UVD_VCPU_CNTL__CLK_EN_MASK); + + /* disable master interrupt */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_MASTINT_EN), 0, + ~UVD_MASTINT_EN__VCPU_EN_MASK); + + /* enable LMI MC and UMC channels */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_LMI_CTRL2), 0, + ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); + + tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); + tmp &= ~UVD_SOFT_RESET__LMI_SOFT_RESET_MASK; + tmp &= ~UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK; + WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); + + /* setup regUVD_LMI_CTRL */ + tmp = RREG32_SOC15(VCN, i, regUVD_LMI_CTRL); + WREG32_SOC15(VCN, i, regUVD_LMI_CTRL, tmp | + UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | + UVD_LMI_CTRL__MASK_MC_URGENT_MASK | + UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK | + UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK); + + /* setup regUVD_MPC_CNTL */ + tmp = RREG32_SOC15(VCN, i, regUVD_MPC_CNTL); + tmp &= ~UVD_MPC_CNTL__REPLACEMENT_MODE_MASK; + tmp |= 0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT; + WREG32_SOC15(VCN, i, regUVD_MPC_CNTL, tmp); + + /* setup UVD_MPC_SET_MUXA0 */ + WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUXA0, + ((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) | + (0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) | + (0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) | + (0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT))); + + /* setup UVD_MPC_SET_MUXB0 */ + WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUXB0, + ((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) | + (0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) | + (0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) | + (0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT))); + + /* setup UVD_MPC_SET_MUX */ + WREG32_SOC15(VCN, i, regUVD_MPC_SET_MUX, + ((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) | + (0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) | + (0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT))); + + vcn_v4_0_5_mc_resume(adev, i); + + /* VCN global tiling registers */ + WREG32_SOC15(VCN, i, regUVD_GFX10_ADDR_CONFIG, + adev->gfx.config.gb_addr_config); + + /* unblock VCPU register access */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_RB_ARB_CTRL), 0, + ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); + + /* release VCPU reset to boot */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, + ~UVD_VCPU_CNTL__BLK_RST_MASK); + + for (j = 0; j < 10; ++j) { + uint32_t status; + + for (k = 0; k < 100; ++k) { + status = RREG32_SOC15(VCN, i, regUVD_STATUS); + if (status & 2) + break; + mdelay(10); + if (amdgpu_emu_mode == 1) + msleep(1); + } + + if (amdgpu_emu_mode == 1) { + r = -1; + if (status & 2) { r = 0; - if (status & 2) - break; - - dev_err(adev->dev, - "VCN[%d] is not responding, trying to reset VCPU!!!\n", i); - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), - UVD_VCPU_CNTL__BLK_RST_MASK, - ~UVD_VCPU_CNTL__BLK_RST_MASK); - mdelay(10); - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, - ~UVD_VCPU_CNTL__BLK_RST_MASK); - - mdelay(10); - r = -1; + break; } + } else { + r = 0; + if (status & 2) + break; + + dev_err(adev->dev, + "VCN[%d] is not responding, trying to reset VCPU!!!\n", i); + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), + UVD_VCPU_CNTL__BLK_RST_MASK, + ~UVD_VCPU_CNTL__BLK_RST_MASK); + mdelay(10); + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, + ~UVD_VCPU_CNTL__BLK_RST_MASK); + + mdelay(10); + r = -1; } + } - if (r) { - dev_err(adev->dev, "VCN[%d] is not responding, giving up!!!\n", i); - return r; - } + if (r) { + dev_err(adev->dev, "VCN[%d] is not responding, giving up!!!\n", i); + return r; + } - /* enable master interrupt */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_MASTINT_EN), - UVD_MASTINT_EN__VCPU_EN_MASK, - ~UVD_MASTINT_EN__VCPU_EN_MASK); + /* enable master interrupt */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_MASTINT_EN), + UVD_MASTINT_EN__VCPU_EN_MASK, + ~UVD_MASTINT_EN__VCPU_EN_MASK); - /* clear the busy bit of VCN_STATUS */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_STATUS), 0, - ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT)); + /* clear the busy bit of VCN_STATUS */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_STATUS), 0, + ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT)); - ring = &adev->vcn.inst[i].ring_enc[0]; - WREG32_SOC15(VCN, i, regVCN_RB1_DB_CTRL, - ring->doorbell_index << VCN_RB1_DB_CTRL__OFFSET__SHIFT | - VCN_RB1_DB_CTRL__EN_MASK); - - WREG32_SOC15(VCN, i, regUVD_RB_BASE_LO, ring->gpu_addr); - WREG32_SOC15(VCN, i, regUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); - WREG32_SOC15(VCN, i, regUVD_RB_SIZE, ring->ring_size / 4); - - tmp = RREG32_SOC15(VCN, i, regVCN_RB_ENABLE); - tmp &= ~(VCN_RB_ENABLE__RB1_EN_MASK); - WREG32_SOC15(VCN, i, regVCN_RB_ENABLE, tmp); - fw_shared->sq.queue_mode |= FW_QUEUE_RING_RESET; - WREG32_SOC15(VCN, i, regUVD_RB_RPTR, 0); - WREG32_SOC15(VCN, i, regUVD_RB_WPTR, 0); - - tmp = RREG32_SOC15(VCN, i, regUVD_RB_RPTR); - WREG32_SOC15(VCN, i, regUVD_RB_WPTR, tmp); - ring->wptr = RREG32_SOC15(VCN, i, regUVD_RB_WPTR); - - tmp = RREG32_SOC15(VCN, i, regVCN_RB_ENABLE); - tmp |= VCN_RB_ENABLE__RB1_EN_MASK; - WREG32_SOC15(VCN, i, regVCN_RB_ENABLE, tmp); - fw_shared->sq.queue_mode &= ~(FW_QUEUE_RING_RESET | FW_QUEUE_DPG_HOLD_OFF); - } + ring = &adev->vcn.inst[i].ring_enc[0]; + WREG32_SOC15(VCN, i, regVCN_RB1_DB_CTRL, + ring->doorbell_index << VCN_RB1_DB_CTRL__OFFSET__SHIFT | + VCN_RB1_DB_CTRL__EN_MASK); + + WREG32_SOC15(VCN, i, regUVD_RB_BASE_LO, ring->gpu_addr); + WREG32_SOC15(VCN, i, regUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(VCN, i, regUVD_RB_SIZE, ring->ring_size / 4); + + tmp = RREG32_SOC15(VCN, i, regVCN_RB_ENABLE); + tmp &= ~(VCN_RB_ENABLE__RB1_EN_MASK); + WREG32_SOC15(VCN, i, regVCN_RB_ENABLE, tmp); + fw_shared->sq.queue_mode |= FW_QUEUE_RING_RESET; + WREG32_SOC15(VCN, i, regUVD_RB_RPTR, 0); + WREG32_SOC15(VCN, i, regUVD_RB_WPTR, 0); + + tmp = RREG32_SOC15(VCN, i, regUVD_RB_RPTR); + WREG32_SOC15(VCN, i, regUVD_RB_WPTR, tmp); + ring->wptr = RREG32_SOC15(VCN, i, regUVD_RB_WPTR); + + tmp = RREG32_SOC15(VCN, i, regVCN_RB_ENABLE); + tmp |= VCN_RB_ENABLE__RB1_EN_MASK; + WREG32_SOC15(VCN, i, regVCN_RB_ENABLE, tmp); + fw_shared->sq.queue_mode &= ~(FW_QUEUE_RING_RESET | FW_QUEUE_DPG_HOLD_OFF); return 0; } @@ -1204,88 +1199,87 @@ static void vcn_v4_0_5_stop_dpg_mode(struct amdgpu_device *adev, int inst_idx) * vcn_v4_0_5_stop - VCN stop * * @adev: amdgpu_device pointer + * @i: instance to stop * * Stop VCN block */ -static int vcn_v4_0_5_stop(struct amdgpu_device *adev) +static int vcn_v4_0_5_stop(struct amdgpu_device *adev, int i) { volatile struct amdgpu_vcn4_fw_shared *fw_shared; uint32_t tmp; - int i, r = 0; + int r = 0; - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { - if (adev->vcn.harvest_config & (1 << i)) - continue; - - fw_shared = adev->vcn.inst[i].fw_shared.cpu_addr; - fw_shared->sq.queue_mode |= FW_QUEUE_DPG_HOLD_OFF; + if (adev->vcn.harvest_config & (1 << i)) + return 0; - if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) { - vcn_v4_0_5_stop_dpg_mode(adev, i); - continue; - } + fw_shared = adev->vcn.inst[i].fw_shared.cpu_addr; + fw_shared->sq.queue_mode |= FW_QUEUE_DPG_HOLD_OFF; - /* wait for vcn idle */ - r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_STATUS, UVD_STATUS__IDLE, 0x7); - if (r) - return r; + if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) { + vcn_v4_0_5_stop_dpg_mode(adev, i); + r = 0; + goto done; + } - tmp = UVD_LMI_STATUS__VCPU_LMI_WRITE_CLEAN_MASK | - UVD_LMI_STATUS__READ_CLEAN_MASK | - UVD_LMI_STATUS__WRITE_CLEAN_MASK | - UVD_LMI_STATUS__WRITE_CLEAN_RAW_MASK; - r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_LMI_STATUS, tmp, tmp); - if (r) - return r; + /* wait for vcn idle */ + r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_STATUS, UVD_STATUS__IDLE, 0x7); + if (r) + goto done; - /* disable LMI UMC channel */ - tmp = RREG32_SOC15(VCN, i, regUVD_LMI_CTRL2); - tmp |= UVD_LMI_CTRL2__STALL_ARB_UMC_MASK; - WREG32_SOC15(VCN, i, regUVD_LMI_CTRL2, tmp); - tmp = UVD_LMI_STATUS__UMC_READ_CLEAN_RAW_MASK | - UVD_LMI_STATUS__UMC_WRITE_CLEAN_RAW_MASK; - r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_LMI_STATUS, tmp, tmp); - if (r) - return r; + tmp = UVD_LMI_STATUS__VCPU_LMI_WRITE_CLEAN_MASK | + UVD_LMI_STATUS__READ_CLEAN_MASK | + UVD_LMI_STATUS__WRITE_CLEAN_MASK | + UVD_LMI_STATUS__WRITE_CLEAN_RAW_MASK; + r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_LMI_STATUS, tmp, tmp); + if (r) + goto done; + + /* disable LMI UMC channel */ + tmp = RREG32_SOC15(VCN, i, regUVD_LMI_CTRL2); + tmp |= UVD_LMI_CTRL2__STALL_ARB_UMC_MASK; + WREG32_SOC15(VCN, i, regUVD_LMI_CTRL2, tmp); + tmp = UVD_LMI_STATUS__UMC_READ_CLEAN_RAW_MASK | + UVD_LMI_STATUS__UMC_WRITE_CLEAN_RAW_MASK; + r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_LMI_STATUS, tmp, tmp); + if (r) + goto done; - /* block VCPU register access */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_RB_ARB_CTRL), - UVD_RB_ARB_CTRL__VCPU_DIS_MASK, - ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); + /* block VCPU register access */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_RB_ARB_CTRL), + UVD_RB_ARB_CTRL__VCPU_DIS_MASK, + ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); - /* reset VCPU */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), - UVD_VCPU_CNTL__BLK_RST_MASK, - ~UVD_VCPU_CNTL__BLK_RST_MASK); + /* reset VCPU */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), + UVD_VCPU_CNTL__BLK_RST_MASK, + ~UVD_VCPU_CNTL__BLK_RST_MASK); - /* disable VCPU clock */ - WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, - ~(UVD_VCPU_CNTL__CLK_EN_MASK)); + /* disable VCPU clock */ + WREG32_P(SOC15_REG_OFFSET(VCN, i, regUVD_VCPU_CNTL), 0, + ~(UVD_VCPU_CNTL__CLK_EN_MASK)); - /* apply soft reset */ - tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); - tmp |= UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK; - WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); - tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); - tmp |= UVD_SOFT_RESET__LMI_SOFT_RESET_MASK; - WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); + /* apply soft reset */ + tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); + tmp |= UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK; + WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); + tmp = RREG32_SOC15(VCN, i, regUVD_SOFT_RESET); + tmp |= UVD_SOFT_RESET__LMI_SOFT_RESET_MASK; + WREG32_SOC15(VCN, i, regUVD_SOFT_RESET, tmp); - /* clear status */ - WREG32_SOC15(VCN, i, regUVD_STATUS, 0); + /* clear status */ + WREG32_SOC15(VCN, i, regUVD_STATUS, 0); - /* apply HW clock gating */ - vcn_v4_0_5_enable_clock_gating(adev, i); + /* apply HW clock gating */ + vcn_v4_0_5_enable_clock_gating(adev, i); - /* enable VCN power gating */ - vcn_v4_0_5_enable_static_power_gating(adev, i); - } + /* enable VCN power gating */ + vcn_v4_0_5_enable_static_power_gating(adev, i); - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { - if (adev->pm.dpm_enabled) - amdgpu_dpm_enable_vcn(adev, false, i); - } +done: + if (adev->pm.dpm_enabled) + amdgpu_dpm_enable_vcn(adev, false, i); - return 0; + return r; } /** @@ -1537,15 +1531,17 @@ static int vcn_v4_0_5_set_powergating_state(struct amdgpu_ip_block *ip_block, enum amd_powergating_state state) { struct amdgpu_device *adev = ip_block->adev; - int ret; + int ret = 0, i; if (state == adev->vcn.cur_state) return 0; - if (state == AMD_PG_STATE_GATE) - ret = vcn_v4_0_5_stop(adev); - else - ret = vcn_v4_0_5_start(adev); + for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { + if (state == AMD_PG_STATE_GATE) + ret |= vcn_v4_0_5_stop(adev, i); + else + ret |= vcn_v4_0_5_start(adev, i); + } if (!ret) adev->vcn.cur_state = state; -- 2.49.0

1 month, 2 weeks

2
3
0 0

Regression between 6.1.135 and 6.1.139, possibly related to ITS mitigations

by Moritz Mühlenhoff

Hi, I'd like to report a regression which seems related to the latest ITS mitigations in Linux 6.1.x: The server in question is a Supermicro SYS-120C-TN10R with a "Intel(R) Xeon(R) Silver 4310 CPU @ 2.10GHz" CPU, running Debian Bookworm. The full output of /proc/cpuinfo is attached as cpuinfo.txt In addition to the kernel changes between 6.1.135 and 6.1.139 there is also some additional invariant, namely the Intel microcode loaded at early boot: On Linux 6.1.135 every works fine with both the 20250211 and 20250512 microcode releases (kern.log is attached as 6.1.135-feb-microcode.log and 6.1.135-may-microcode.log) With 6.1.139 and the February microcode, oopses appear related to clear_bhb_loop() (which may be related to "x86/its: Align RETs in BHB clear sequence to avoid thunking"?). This is captured in 6.1.139-feb-microcode.log. With 6.1.139 and the May microcode, the system mostly crashes on bootup (in my tests it crashed in three out of four attempts). I've captured both the crash (6.1.139-may-microcode-crash.log) and a working boot (6.1.139-may-microcode-noncrash.log). If you need any additional information, please let me know! Cheers, Moritz

1 month, 2 weeks

3
3
0 0

[PATCH v4 rc] iommu: Skip PASID validation for devices without PASID capability

by Tushar Dave

Generally PASID support requires ACS settings that usually create single device groups, but there are some niche cases where we can get multi-device groups and still have working PASID support. The primary issue is that PCI switches are not required to treat PASID tagged TLPs specially so appropriate ACS settings are required to route all TLPs to the host bridge if PASID is going to work properly. pci_enable_pasid() does check that each device that will use PASID has the proper ACS settings to achieve this routing. However, no-PASID devices can be combined with PASID capable devices within the same topology using non-uniform ACS settings. In this case the no-PASID devices may not have strict route to host ACS flags and end up being grouped with the PASID devices. This configuration fails to allow use of the PASID within the iommu core code which wrongly checks if the no-PASID device supports PASID. Fix this by ignoring no-PASID devices during the PASID validation. They will never issue a PASID TLP anyhow so they can be ignored. Fixes: c404f55c26fc ("iommu: Validate the PASID in iommu_attach_device_pasid()") Cc: stable(a)vger.kernel.org Signed-off-by: Tushar Dave <tdave(a)nvidia.com> --- changes in v4: - rebase to 6.15-rc7 drivers/iommu/iommu.c | 43 ++++++++++++++++++++++++++++--------------- 1 file changed, 28 insertions(+), 15 deletions(-) diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c index 4f91a740c15f..9d728800a862 100644 --- a/drivers/iommu/iommu.c +++ b/drivers/iommu/iommu.c @@ -3366,10 +3366,12 @@ static int __iommu_set_group_pasid(struct iommu_domain *domain, int ret; for_each_group_device(group, device) { - ret = domain->ops->set_dev_pasid(domain, device->dev, - pasid, old); - if (ret) - goto err_revert; + if (device->dev->iommu->max_pasids > 0) { + ret = domain->ops->set_dev_pasid(domain, device->dev, + pasid, old); + if (ret) + goto err_revert; + } } return 0; @@ -3379,15 +3381,18 @@ static int __iommu_set_group_pasid(struct iommu_domain *domain, for_each_group_device(group, device) { if (device == last_gdev) break; - /* - * If no old domain, undo the succeeded devices/pasid. - * Otherwise, rollback the succeeded devices/pasid to the old - * domain. And it is a driver bug to fail attaching with a - * previously good domain. - */ - if (!old || WARN_ON(old->ops->set_dev_pasid(old, device->dev, + if (device->dev->iommu->max_pasids > 0) { + /* + * If no old domain, undo the succeeded devices/pasid. + * Otherwise, rollback the succeeded devices/pasid to + * the old domain. And it is a driver bug to fail + * attaching with a previously good domain. + */ + if (!old || + WARN_ON(old->ops->set_dev_pasid(old, device->dev, pasid, domain))) - iommu_remove_dev_pasid(device->dev, pasid, domain); + iommu_remove_dev_pasid(device->dev, pasid, domain); + } } return ret; } @@ -3398,8 +3403,10 @@ static void __iommu_remove_group_pasid(struct iommu_group *group, { struct group_device *device; - for_each_group_device(group, device) - iommu_remove_dev_pasid(device->dev, pasid, domain); + for_each_group_device(group, device) { + if (device->dev->iommu->max_pasids > 0) + iommu_remove_dev_pasid(device->dev, pasid, domain); + } } /* @@ -3440,7 +3447,13 @@ int iommu_attach_device_pasid(struct iommu_domain *domain, mutex_lock(&group->mutex); for_each_group_device(group, device) { - if (pasid >= device->dev->iommu->max_pasids) { + /* + * Skip PASID validation for devices without PASID support + * (max_pasids = 0). These devices cannot issue transactions + * with PASID, so they don't affect group's PASID usage. + */ + if ((device->dev->iommu->max_pasids > 0) && + (pasid >= device->dev->iommu->max_pasids)) { ret = -EINVAL; goto out_unlock; } -- 2.34.1

1 month, 2 weeks

4
3
0 0

6.12.29: amdgpu-related "fail"-messages during boot; bisected

by Rainer Fiebig

Hi! After updating to linux-6.12.29, I see lots of "fail"-messages during boot: May 19 23:39:09 LUX kernel: [ 4.819552] amdgpu 0000:30:00.0: amdgpu: [drm] amdgpu: DP AUX transfer fail:4 Bisecting for drivers/gpu/drm/amd had this result: > git bisect bad 2d63e66f7ba7b88b87e72155a33b970c81cf4664 is the first bad commit commit 2d63e66f7ba7b88b87e72155a33b970c81cf4664 (HEAD) Author: Wayne Lin <Wayne.Lin(a)amd.com> Date: Sun Apr 20 19:22:14 2025 +0800 drm/amd/display: Fix wrong handling for AUX_DEFER case commit 65924ec69b29296845c7f628112353438e63ea56 upstream. The system (Ryzen 3 5600G, latest BIOS) is stable so far but the error-messages are not nice to see. Thanks. Rainer Fiebig -- The truth always turns out to be simpler than you thought. Richard Feynman

1 month, 2 weeks

3
3
0 0

[PATCH v2] mm/hugetlb: fix a deadlock with pagecache_folio and hugetlb_fault_mutex_table

by Gavin Guo

The patch fixes a deadlock which can be triggered by an internal syzkaller [1] reproducer and captured by bpftrace script [2] and its log [3] in this scenario: Process 1 Process 2 --- --- hugetlb_fault mutex_lock(B) // take B filemap_lock_hugetlb_folio filemap_lock_folio __filemap_get_folio folio_lock(A) // take A hugetlb_wp mutex_unlock(B) // release B ... hugetlb_fault ... mutex_lock(B) // take B filemap_lock_hugetlb_folio filemap_lock_folio __filemap_get_folio folio_lock(A) // blocked unmap_ref_private ... mutex_lock(B) // retake and blocked This is a ABBA deadlock involving two locks: - Lock A: pagecache_folio lock - Lock B: hugetlb_fault_mutex_table lock The deadlock occurs between two processes as follows: 1. The first process (let’s call it Process 1) is handling a copy-on-write (COW) operation on a hugepage via hugetlb_wp. Due to insufficient reserved hugetlb pages, Process 1, owner of the reserved hugetlb page, attempts to unmap a hugepage owned by another process (non-owner) to satisfy the reservation. Before unmapping, Process 1 acquires lock B (hugetlb_fault_mutex_table lock) and then lock A (pagecache_folio lock). To proceed with the unmap, it releases Lock B but retains Lock A. After the unmap, Process 1 tries to reacquire Lock B. However, at this point, Lock B has already been acquired by another process. 2. The second process (Process 2) enters the hugetlb_fault handler during the unmap operation. It successfully acquires Lock B (hugetlb_fault_mutex_table lock) that was just released by Process 1, but then attempts to acquire Lock A (pagecache_folio lock), which is still held by Process 1. As a result, Process 1 (holding Lock A) is blocked waiting for Lock B (held by Process 2), while Process 2 (holding Lock B) is blocked waiting for Lock A (held by Process 1), constructing a ABBA deadlock scenario. The error message: INFO: task repro_20250402_:13229 blocked for more than 64 seconds. Not tainted 6.15.0-rc3+ #24 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:repro_20250402_ state:D stack:25856 pid:13229 tgid:13228 ppid:3513 task_flags:0x400040 flags:0x00004006 Call Trace: <TASK> __schedule+0x1755/0x4f50 schedule+0x158/0x330 schedule_preempt_disabled+0x15/0x30 __mutex_lock+0x75f/0xeb0 hugetlb_wp+0xf88/0x3440 hugetlb_fault+0x14c8/0x2c30 trace_clock_x86_tsc+0x20/0x20 do_user_addr_fault+0x61d/0x1490 exc_page_fault+0x64/0x100 asm_exc_page_fault+0x26/0x30 RIP: 0010:__put_user_4+0xd/0x20 copy_process+0x1f4a/0x3d60 kernel_clone+0x210/0x8f0 __x64_sys_clone+0x18d/0x1f0 do_syscall_64+0x6a/0x120 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x41b26d </TASK> INFO: task repro_20250402_:13229 is blocked on a mutex likely owned by task repro_20250402_:13250. task:repro_20250402_ state:D stack:28288 pid:13250 tgid:13228 ppid:3513 task_flags:0x400040 flags:0x00000006 Call Trace: <TASK> __schedule+0x1755/0x4f50 schedule+0x158/0x330 io_schedule+0x92/0x110 folio_wait_bit_common+0x69a/0xba0 __filemap_get_folio+0x154/0xb70 hugetlb_fault+0xa50/0x2c30 trace_clock_x86_tsc+0x20/0x20 do_user_addr_fault+0xace/0x1490 exc_page_fault+0x64/0x100 asm_exc_page_fault+0x26/0x30 RIP: 0033:0x402619 </TASK> INFO: task repro_20250402_:13250 blocked for more than 65 seconds. Not tainted 6.15.0-rc3+ #24 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:repro_20250402_ state:D stack:28288 pid:13250 tgid:13228 ppid:3513 task_flags:0x400040 flags:0x00000006 Call Trace: <TASK> __schedule+0x1755/0x4f50 schedule+0x158/0x330 io_schedule+0x92/0x110 folio_wait_bit_common+0x69a/0xba0 __filemap_get_folio+0x154/0xb70 hugetlb_fault+0xa50/0x2c30 trace_clock_x86_tsc+0x20/0x20 do_user_addr_fault+0xace/0x1490 exc_page_fault+0x64/0x100 asm_exc_page_fault+0x26/0x30 RIP: 0033:0x402619 </TASK> Showing all locks held in the system: 1 lock held by khungtaskd/35: #0: ffffffff879a7440 (rcu_read_lock){....}-{1:3}, at: debug_show_all_locks+0x30/0x180 2 locks held by repro_20250402_/13229: #0: ffff888017d801e0 (&mm->mmap_lock){++++}-{4:4}, at: lock_mm_and_find_vma+0x37/0x300 #1: ffff888000fec848 (&hugetlb_fault_mutex_table[i]){+.+.}-{4:4}, at: hugetlb_wp+0xf88/0x3440 3 locks held by repro_20250402_/13250: #0: ffff8880177f3d08 (vm_lock){++++}-{0:0}, at: do_user_addr_fault+0x41b/0x1490 #1: ffff888000fec848 (&hugetlb_fault_mutex_table[i]){+.+.}-{4:4}, at: hugetlb_fault+0x3b8/0x2c30 #2: ffff8880129500e8 (&resv_map->rw_sema){++++}-{4:4}, at: hugetlb_fault+0x494/0x2c30 Link: https://drive.google.com/file/d/1DVRnIW-vSayU5J1re9Ct_br3jJQU6Vpb/view?usp=… [1] Link: https://github.com/bboymimi/bpftracer/blob/master/scripts/hugetlb_lock_debu… [2] Link: https://drive.google.com/file/d/1bWq2-8o-BJAuhoHWX7zAhI6ggfhVzQUI/view?usp=… [3] Fixes: 40549ba8f8e0 ("hugetlb: use new vma_lock for pmd sharing synchronization") Cc: stable(a)vger.kernel.org Cc: Hugh Dickins <hughd(a)google.com> Cc: Florent Revest <revest(a)google.com> Cc: Gavin Shan <gshan(a)redhat.com> Suggested-by: Oscar Salvador <osalvador(a)suse.de> Signed-off-by: Gavin Guo <gavinguo(a)igalia.com> --- V1 -> V2 Suggested-by Oscar Salvador: - Use folio_test_locked to replace the unnecessary parameter passing. mm/hugetlb.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 7ae38bfb9096..ed501f134eff 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -6226,6 +6226,12 @@ static vm_fault_t hugetlb_wp(struct folio *pagecache_folio, u32 hash; folio_put(old_folio); + /* + * The pagecache_folio needs to be unlocked to avoid + * deadlock when the child unmaps the folio. + */ + if (pagecache_folio) + folio_unlock(pagecache_folio); /* * Drop hugetlb_fault_mutex and vma_lock before * unmapping. unmapping needs to hold vma_lock @@ -6823,8 +6829,13 @@ vm_fault_t hugetlb_fault(struct mm_struct *mm, struct vm_area_struct *vma, out_ptl: spin_unlock(vmf.ptl); + /* + * hugetlb_wp() might have already unlocked pagecache_folio, so + * skip it if that is the case. + */ if (pagecache_folio) { - folio_unlock(pagecache_folio); + if (folio_test_locked(pagecache_folio)) + folio_unlock(pagecache_folio); folio_put(pagecache_folio); } out_mutex: base-commit: 4a95bc121ccdaee04c4d72f84dbfa6b880a514b6 -- 2.43.0

1 month, 2 weeks

3
6
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror May 2025