* [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails
@ 2021-08-19 1:53 Wenpeng Liang
2021-08-19 1:53 ` [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset Wenpeng Liang
2021-08-19 2:16 ` [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
0 siblings, 2 replies; 4+ messages in thread
From: Wenpeng Liang @ 2021-08-19 1:53 UTC (permalink / raw)
To: dledford, jgg; +Cc: linux-rdma, linuxarm, liangwenpeng
From: Yangyang Li <liyangyang20@huawei.com>
If rereg_user_mr fails, mr needs to stay in a safe state. The mr state is
backed up when entering the rereg_user_mr function, and the mr state is
restored when the rereg fails.
Fixes: 4e9fc1dae2a9 ("RDMA/hns: Optimize the MR registration process")
Reported-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Yangyang Li <liyangyang20@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
---
drivers/infiniband/hw/hns/hns_roce_mr.c | 28 ++++++++++++++++++++++++++++
1 file changed, 28 insertions(+)
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index 006c84b..d0870b4 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -285,6 +285,27 @@ struct ib_mr *hns_roce_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
return ERR_PTR(ret);
}
+static void copy_mr(struct hns_roce_mr *dst, struct hns_roce_mr *src)
+{
+ dst->enabled = src->enabled;
+ dst->iova = src->iova;
+ dst->size = src->size;
+ dst->pd = src->pd;
+ dst->access = src->access;
+}
+
+static void store_rereg_mr(struct hns_roce_mr *mr_bak,
+ struct hns_roce_mr *mr)
+{
+ copy_mr(mr_bak, mr);
+}
+
+static void restore_rereg_mr(struct hns_roce_mr *mr_bak,
+ struct hns_roce_mr *mr)
+{
+ copy_mr(mr, mr_bak);
+}
+
struct ib_mr *hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start,
u64 length, u64 virt_addr,
int mr_access_flags, struct ib_pd *pd,
@@ -294,9 +315,12 @@ struct ib_mr *hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start,
struct ib_device *ib_dev = &hr_dev->ib_dev;
struct hns_roce_mr *mr = to_hr_mr(ibmr);
struct hns_roce_cmd_mailbox *mailbox;
+ struct hns_roce_mr mr_bak;
unsigned long mtpt_idx;
int ret;
+ store_rereg_mr(&mr_bak, mr);
+
if (!mr->enabled)
return ERR_PTR(-EINVAL);
@@ -349,7 +373,11 @@ struct ib_mr *hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start,
mr->enabled = 1;
+ hns_roce_free_cmd_mailbox(hr_dev, mailbox);
+ return NULL;
+
free_cmd_mbox:
+ restore_rereg_mr(&mr_bak, mr);
hns_roce_free_cmd_mailbox(hr_dev, mailbox);
return ERR_PTR(ret);
--
2.8.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset
2021-08-19 1:53 [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
@ 2021-08-19 1:53 ` Wenpeng Liang
2021-08-19 2:16 ` Wenpeng Liang
2021-08-19 2:16 ` [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
1 sibling, 1 reply; 4+ messages in thread
From: Wenpeng Liang @ 2021-08-19 1:53 UTC (permalink / raw)
To: dledford, jgg; +Cc: linux-rdma, linuxarm, liangwenpeng
From: Lang Cheng <chenglang@huawei.com>
During the reset, the driver calls dma_pool_destroy() to release the
dma_pool resources. If the dma_pool_free interface is called during the
modify_qp operation, an exception will occur.
[15834.440744] Unable to handle kernel paging request at virtual address
ffffa2cfc7725678
...
[15834.660596] Call trace:
[15834.663033] queued_spin_lock_slowpath+0x224/0x308
[15834.667802] _raw_spin_lock_irqsave+0x78/0x88
[15834.672140] dma_pool_free+0x34/0x118
[15834.675799] hns_roce_free_cmd_mailbox+0x54/0x88 [hns_roce_hw_v2]
[15834.681872] hns_roce_v2_qp_modify.isra.57+0xcc/0x120 [hns_roce_hw_v2]
[15834.688376] hns_roce_v2_modify_qp+0x4d4/0x1ef8 [hns_roce_hw_v2]
[15834.694362] hns_roce_modify_qp+0x214/0x5a8 [hns_roce_hw_v2]
[15834.699996] _ib_modify_qp+0xf0/0x308
[15834.703642] ib_modify_qp+0x38/0x48
[15834.707118] rt_ktest_modify_qp+0x14c/0x998 [rdma_test]
...
[15837.269216] Unable to handle kernel paging request at virtual address
000197c995a1d1b4
...
[15837.480898] Call trace:
[15837.483335] __free_pages+0x28/0x78
[15837.486807] dma_direct_free_pages+0xa0/0xe8
[15837.491058] dma_direct_free+0x48/0x60
[15837.494790] dma_free_attrs+0xa4/0xe8
[15837.498449] hns_roce_buf_free+0xb0/0x150 [hns_roce_hw_v2]
[15837.503918] mtr_free_bufs.isra.1+0x88/0xc0 [hns_roce_hw_v2]
[15837.509558] hns_roce_mtr_destroy+0x60/0x80 [hns_roce_hw_v2]
[15837.515198] hns_roce_v2_cleanup_eq_table+0x1d0/0x2a0 [hns_roce_hw_v2]
[15837.521701] hns_roce_exit+0x108/0x1e0 [hns_roce_hw_v2]
[15837.526908] __hns_roce_hw_v2_uninit_instance.isra.75+0x70/0xb8 [hns_roce_hw_v2]
[15837.534276] hns_roce_hw_v2_uninit_instance+0x64/0x80 [hns_roce_hw_v2]
[15837.540786] hclge_uninit_client_instance+0xe8/0x1e8 [hclge]
[15837.546419] hnae3_uninit_client_instance+0xc4/0x118 [hnae3]
[15837.552052] hnae3_unregister_client+0x16c/0x1f0 [hnae3]
[15837.557346] hns_roce_hw_v2_exit+0x34/0x50 [hns_roce_hw_v2]
[15837.562895] __arm64_sys_delete_module+0x208/0x268
[15837.567665] el0_svc_common.constprop.4+0x110/0x200
[15837.572520] do_el0_svc+0x34/0x98
[15837.575821] el0_svc+0x14/0x40
[15837.578862] el0_sync_handler+0xb0/0x2d0
[15837.582766] el0_sync+0x140/0x180
It is caused by two concurrent processes:
uninit_instance->dma_pool_destroy(cmdq)
modify_qp->dma_poll_free(cmdq)
Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
Signed-off-by: Lang Cheng <chenglang@huawei.com>
Signed-off-by: Jiaran Zhang <zhangjiaran@huawei.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
---
drivers/infiniband/hw/hns/hns_roce_cmd.c | 12 ++++++++++++
drivers/infiniband/hw/hns/hns_roce_device.h | 1 +
2 files changed, 13 insertions(+)
diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.c b/drivers/infiniband/hw/hns/hns_roce_cmd.c
index 8f68cc3..3dfb97a 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cmd.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cmd.c
@@ -198,12 +198,16 @@ int hns_roce_cmd_init(struct hns_roce_dev *hr_dev)
if (!hr_dev->cmd.pool)
return -ENOMEM;
+ init_rwsem(&hr_dev->cmd.mb_rwsem);
+
return 0;
}
void hns_roce_cmd_cleanup(struct hns_roce_dev *hr_dev)
{
+ down_write(&hr_dev->cmd.mb_rwsem);
dma_pool_destroy(hr_dev->cmd.pool);
+ up_write(&hr_dev->cmd.mb_rwsem);
}
int hns_roce_cmd_use_events(struct hns_roce_dev *hr_dev)
@@ -237,8 +241,10 @@ void hns_roce_cmd_use_polling(struct hns_roce_dev *hr_dev)
{
struct hns_roce_cmdq *hr_cmd = &hr_dev->cmd;
+ down_write(&hr_dev->cmd.mb_rwsem);
kfree(hr_cmd->context);
hr_cmd->use_events = 0;
+ up_write(&hr_dev->cmd.mb_rwsem);
up(&hr_cmd->poll_sem);
}
@@ -252,9 +258,12 @@ hns_roce_alloc_cmd_mailbox(struct hns_roce_dev *hr_dev)
if (!mailbox)
return ERR_PTR(-ENOMEM);
+ down_read(&hr_dev->cmd.mb_rwsem);
mailbox->buf =
dma_pool_alloc(hr_dev->cmd.pool, GFP_KERNEL, &mailbox->dma);
if (!mailbox->buf) {
+ up_read(&hr_dev->cmd.mb_rwsem);
+
kfree(mailbox);
return ERR_PTR(-ENOMEM);
}
@@ -269,5 +278,8 @@ void hns_roce_free_cmd_mailbox(struct hns_roce_dev *hr_dev,
return;
dma_pool_free(hr_dev->cmd.pool, mailbox->buf, mailbox->dma);
+
+ up_read(&hr_dev->cmd.mb_rwsem);
+
kfree(mailbox);
}
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 0c3eb11..90d8ef8 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -571,6 +571,7 @@ struct hns_roce_cmdq {
* close device, switch into poll mode(non event mode)
*/
u8 use_events;
+ struct rw_semaphore mb_rwsem;
};
struct hns_roce_cmd_mailbox {
--
2.8.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails
2021-08-19 1:53 [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
2021-08-19 1:53 ` [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset Wenpeng Liang
@ 2021-08-19 2:16 ` Wenpeng Liang
1 sibling, 0 replies; 4+ messages in thread
From: Wenpeng Liang @ 2021-08-19 2:16 UTC (permalink / raw)
To: dledford, jgg; +Cc: linux-rdma, linuxarm
Please ignore this patch.
I mistakenly sent two unrelated patches together,
and I will send the two patches separately
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset
2021-08-19 1:53 ` [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset Wenpeng Liang
@ 2021-08-19 2:16 ` Wenpeng Liang
0 siblings, 0 replies; 4+ messages in thread
From: Wenpeng Liang @ 2021-08-19 2:16 UTC (permalink / raw)
To: dledford, jgg; +Cc: linux-rdma, linuxarm
Please ignore this patch.
I mistakenly sent two unrelated patches together,
and I will send the two patches separately.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2021-08-19 2:16 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-08-19 1:53 [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
2021-08-19 1:53 ` [PATCH v2 for-next] RDMA/hns: Solve the problem that dma_pool is used during the reset Wenpeng Liang
2021-08-19 2:16 ` Wenpeng Liang
2021-08-19 2:16 ` [PATCH for-next] RDMA/hns: Restore mr status when rereg mr fails Wenpeng Liang
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).