* [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw
@ 2017-08-09 1:38 Henry Orosco
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
0 siblings, 1 reply; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f, Henry Orosco
The following patch series has fixes for the i40iw driver.
V2 - Removed 1 patch from series
--------------------------------------------------------
Git Repo to pull from:
https://github.com/horosco/linux.git
tag:
for-v4.13-08-07-2017-v2
Commits of interest:
commit 08bc3c4ee4c759655b2250b59ddec80950772f6a
Date: Wed Jul 5 13:33:54 2017 -0500
i40iw: Fix potential fcn_id_array out of bounds
to
commit a18153f9af05c1f03ee449cc881fd63e9f83d7fc
Date: Mon Jul 3 15:45:32 2017 -0500
i40iw: Fix parsing of query/commit FPM buffers
--------------------------------------------------------
Chien Tin Tung (1):
i40iw: Fix parsing of query/commit FPM buffers
Christopher N Bednarz (2):
i40iw: Use correct alignment for CQ0 memory
i40iw: Fix potential fcn_id_array out of bounds
Mustafa Ismail (2):
i40iw: Correct variable names
i40iw: Fix typecast of tcp_seq_num
Shiraz Saleem (1):
i40iw: Improve CQP timeout logic
drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 134 ++++++++++++++++++++---------
drivers/infiniband/hw/i40iw/i40iw_d.h | 4 +-
drivers/infiniband/hw/i40iw/i40iw_p.h | 14 +--
drivers/infiniband/hw/i40iw/i40iw_puda.c | 2 +-
drivers/infiniband/hw/i40iw/i40iw_status.h | 2 +-
drivers/infiniband/hw/i40iw/i40iw_type.h | 5 ++
drivers/infiniband/hw/i40iw/i40iw_uk.c | 8 +-
drivers/infiniband/hw/i40iw/i40iw_utils.c | 22 +++--
8 files changed, 129 insertions(+), 62 deletions(-)
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH v2 1/6] i40iw: Fix parsing of query/commit FPM buffers
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-09 1:38 ` [PATCH v2 2/6] i40iw: Correct variable names Henry Orosco
` (5 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f, Chien Tin Tung,
Henry Orosco
From: Chien Tin Tung <chien.tin.tung-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Parsing of commit/query Host Memory Cache Function Private Memory
is not skipping over reserved fields and incorrectly assigning
those values into object's base/cnt/max_cnt fields. Skip over
reserved fields and set correct values. Also correct memory
alignment requirement for commit/query FPM buffers.
Signed-off-by: Chien Tin Tung <chien.tin.tung-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Christopher N Bednarz <christopher.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 121 +++++++++++++++++++++----------
drivers/infiniband/hw/i40iw/i40iw_d.h | 4 +-
2 files changed, 83 insertions(+), 42 deletions(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 9ec1ae9..ef4a73c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -130,20 +130,32 @@ static enum i40iw_status_code i40iw_sc_parse_fpm_commit_buf(
u64 base = 0;
u32 i, j;
u32 k = 0;
- u32 low;
/* copy base values in obj_info */
- for (i = I40IW_HMC_IW_QP, j = 0;
- i <= I40IW_HMC_IW_PBLE; i++, j += 8) {
+ for (i = I40IW_HMC_IW_QP, j = 0; i <= I40IW_HMC_IW_PBLE; i++, j += 8) {
+ if ((i == I40IW_HMC_IW_SRQ) ||
+ (i == I40IW_HMC_IW_FSIMC) ||
+ (i == I40IW_HMC_IW_FSIAV)) {
+ info[i].base = 0;
+ info[i].cnt = 0;
+ continue;
+ }
get_64bit_val(buf, j, &temp);
info[i].base = RS_64_1(temp, 32) * 512;
if (info[i].base > base) {
base = info[i].base;
k = i;
}
- low = (u32)(temp);
- if (low)
- info[i].cnt = low;
+ if (i == I40IW_HMC_IW_APBVT_ENTRY) {
+ info[i].cnt = 1;
+ continue;
+ }
+ if (i == I40IW_HMC_IW_QP)
+ info[i].cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_QPS);
+ else if (i == I40IW_HMC_IW_CQ)
+ info[i].cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_CQS);
+ else
+ info[i].cnt = (u32)(temp);
}
size = info[k].cnt * info[k].size + info[k].base;
if (size & 0x1FFFFF)
@@ -155,6 +167,31 @@ static enum i40iw_status_code i40iw_sc_parse_fpm_commit_buf(
}
/**
+ * i40iw_sc_decode_fpm_query() - Decode a 64 bit value into max count and size
+ * @buf: ptr to fpm query buffer
+ * @buf_idx: index into buf
+ * @info: ptr to i40iw_hmc_obj_info struct
+ * @rsrc_idx: resource index into info
+ *
+ * Decode a 64 bit value from fpm query buffer into max count and size
+ */
+static u64 i40iw_sc_decode_fpm_query(u64 *buf,
+ u32 buf_idx,
+ struct i40iw_hmc_obj_info *obj_info,
+ u32 rsrc_idx)
+{
+ u64 temp;
+ u32 size;
+
+ get_64bit_val(buf, buf_idx, &temp);
+ obj_info[rsrc_idx].max_cnt = (u32)temp;
+ size = (u32)RS_64_1(temp, 32);
+ obj_info[rsrc_idx].size = LS_64_1(1, size);
+
+ return temp;
+}
+
+/**
* i40iw_sc_parse_fpm_query_buf() - parses fpm query buffer
* @buf: ptr to fpm query buffer
* @info: ptr to i40iw_hmc_obj_info struct
@@ -168,9 +205,9 @@ static enum i40iw_status_code i40iw_sc_parse_fpm_query_buf(
struct i40iw_hmc_info *hmc_info,
struct i40iw_hmc_fpm_misc *hmc_fpm_misc)
{
- u64 temp;
struct i40iw_hmc_obj_info *obj_info;
- u32 i, j, size;
+ u64 temp;
+ u32 size;
u16 max_pe_sds;
obj_info = hmc_info->hmc_obj;
@@ -185,41 +222,52 @@ static enum i40iw_status_code i40iw_sc_parse_fpm_query_buf(
hmc_fpm_misc->max_sds = max_pe_sds;
hmc_info->sd_table.sd_cnt = max_pe_sds + hmc_info->first_sd_index;
- for (i = I40IW_HMC_IW_QP, j = 8;
- i <= I40IW_HMC_IW_ARP; i++, j += 8) {
- get_64bit_val(buf, j, &temp);
- if (i == I40IW_HMC_IW_QP)
- obj_info[i].max_cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_QPS);
- else if (i == I40IW_HMC_IW_CQ)
- obj_info[i].max_cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_CQS);
- else
- obj_info[i].max_cnt = (u32)temp;
+ get_64bit_val(buf, 8, &temp);
+ obj_info[I40IW_HMC_IW_QP].max_cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_QPS);
+ size = (u32)RS_64_1(temp, 32);
+ obj_info[I40IW_HMC_IW_QP].size = LS_64_1(1, size);
- size = (u32)RS_64_1(temp, 32);
- obj_info[i].size = ((u64)1 << size);
- }
- for (i = I40IW_HMC_IW_MR, j = 48;
- i <= I40IW_HMC_IW_PBLE; i++, j += 8) {
- get_64bit_val(buf, j, &temp);
- obj_info[i].max_cnt = (u32)temp;
- size = (u32)RS_64_1(temp, 32);
- obj_info[i].size = LS_64_1(1, size);
- }
+ get_64bit_val(buf, 16, &temp);
+ obj_info[I40IW_HMC_IW_CQ].max_cnt = (u32)RS_64(temp, I40IW_QUERY_FPM_MAX_CQS);
+ size = (u32)RS_64_1(temp, 32);
+ obj_info[I40IW_HMC_IW_CQ].size = LS_64_1(1, size);
+
+ i40iw_sc_decode_fpm_query(buf, 32, obj_info, I40IW_HMC_IW_HTE);
+ i40iw_sc_decode_fpm_query(buf, 40, obj_info, I40IW_HMC_IW_ARP);
+
+ obj_info[I40IW_HMC_IW_APBVT_ENTRY].size = 8192;
+ obj_info[I40IW_HMC_IW_APBVT_ENTRY].max_cnt = 1;
+
+ i40iw_sc_decode_fpm_query(buf, 48, obj_info, I40IW_HMC_IW_MR);
+ i40iw_sc_decode_fpm_query(buf, 56, obj_info, I40IW_HMC_IW_XF);
- get_64bit_val(buf, 120, &temp);
- hmc_fpm_misc->max_ceqs = (u8)RS_64(temp, I40IW_QUERY_FPM_MAX_CEQS);
- get_64bit_val(buf, 120, &temp);
- hmc_fpm_misc->ht_multiplier = RS_64(temp, I40IW_QUERY_FPM_HTMULTIPLIER);
- get_64bit_val(buf, 120, &temp);
- hmc_fpm_misc->timer_bucket = RS_64(temp, I40IW_QUERY_FPM_TIMERBUCKET);
get_64bit_val(buf, 64, &temp);
+ obj_info[I40IW_HMC_IW_XFFL].max_cnt = (u32)temp;
+ obj_info[I40IW_HMC_IW_XFFL].size = 4;
hmc_fpm_misc->xf_block_size = RS_64(temp, I40IW_QUERY_FPM_XFBLOCKSIZE);
if (!hmc_fpm_misc->xf_block_size)
return I40IW_ERR_INVALID_SIZE;
+
+ i40iw_sc_decode_fpm_query(buf, 72, obj_info, I40IW_HMC_IW_Q1);
+
get_64bit_val(buf, 80, &temp);
+ obj_info[I40IW_HMC_IW_Q1FL].max_cnt = (u32)temp;
+ obj_info[I40IW_HMC_IW_Q1FL].size = 4;
hmc_fpm_misc->q1_block_size = RS_64(temp, I40IW_QUERY_FPM_Q1BLOCKSIZE);
if (!hmc_fpm_misc->q1_block_size)
return I40IW_ERR_INVALID_SIZE;
+
+ i40iw_sc_decode_fpm_query(buf, 88, obj_info, I40IW_HMC_IW_TIMER);
+
+ get_64bit_val(buf, 112, &temp);
+ obj_info[I40IW_HMC_IW_PBLE].max_cnt = (u32)temp;
+ obj_info[I40IW_HMC_IW_PBLE].size = 8;
+
+ get_64bit_val(buf, 120, &temp);
+ hmc_fpm_misc->max_ceqs = (u8)RS_64(temp, I40IW_QUERY_FPM_MAX_CEQS);
+ hmc_fpm_misc->ht_multiplier = RS_64(temp, I40IW_QUERY_FPM_HTMULTIPLIER);
+ hmc_fpm_misc->timer_bucket = RS_64(temp, I40IW_QUERY_FPM_TIMERBUCKET);
+
return 0;
}
@@ -3392,13 +3440,6 @@ enum i40iw_status_code i40iw_sc_init_iw_hmc(struct i40iw_sc_dev *dev, u8 hmc_fn_
hmc_info->sd_table.sd_entry = virt_mem.va;
}
- /* fill size of objects which are fixed */
- hmc_info->hmc_obj[I40IW_HMC_IW_XFFL].size = 4;
- hmc_info->hmc_obj[I40IW_HMC_IW_Q1FL].size = 4;
- hmc_info->hmc_obj[I40IW_HMC_IW_PBLE].size = 8;
- hmc_info->hmc_obj[I40IW_HMC_IW_APBVT_ENTRY].size = 8192;
- hmc_info->hmc_obj[I40IW_HMC_IW_APBVT_ENTRY].max_cnt = 1;
-
return ret_code;
}
diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h
index a39ac12..2ebaadb 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_d.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_d.h
@@ -1507,8 +1507,8 @@ enum {
I40IW_CQ0_ALIGNMENT_MASK = (256 - 1),
I40IW_HOST_CTX_ALIGNMENT_MASK = (4 - 1),
I40IW_SHADOWAREA_MASK = (128 - 1),
- I40IW_FPM_QUERY_BUF_ALIGNMENT_MASK = 0,
- I40IW_FPM_COMMIT_BUF_ALIGNMENT_MASK = 0
+ I40IW_FPM_QUERY_BUF_ALIGNMENT_MASK = (4 - 1),
+ I40IW_FPM_COMMIT_BUF_ALIGNMENT_MASK = (4 - 1)
};
enum i40iw_alignment {
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v2 2/6] i40iw: Correct variable names
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2017-08-09 1:38 ` [PATCH v2 1/6] i40iw: Fix parsing of query/commit FPM buffers Henry Orosco
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-09 1:38 ` [PATCH v2 3/6] i40iw: Improve CQP timeout logic Henry Orosco
` (4 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f, Mustafa Ismail,
Henry Orosco
From: Mustafa Ismail <mustafa.ismail-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Fix incorrect naming of status code and struct. Use inline
instead of immediate.
Signed-off-by: Mustafa Ismail <mustafa.ismail-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_status.h | 2 +-
drivers/infiniband/hw/i40iw/i40iw_uk.c | 6 +++---
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_status.h b/drivers/infiniband/hw/i40iw/i40iw_status.h
index 91c4217..f7013f1 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_status.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_status.h
@@ -62,7 +62,7 @@ enum i40iw_status_code {
I40IW_ERR_INVALID_ALIGNMENT = -23,
I40IW_ERR_FLUSHED_QUEUE = -24,
I40IW_ERR_INVALID_PUSH_PAGE_INDEX = -25,
- I40IW_ERR_INVALID_IMM_DATA_SIZE = -26,
+ I40IW_ERR_INVALID_INLINE_DATA_SIZE = -26,
I40IW_ERR_TIMEOUT = -27,
I40IW_ERR_OPCODE_MISMATCH = -28,
I40IW_ERR_CQP_COMPL_ERROR = -29,
diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index b0d3a0e..70a6b41 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -435,7 +435,7 @@ static enum i40iw_status_code i40iw_inline_rdma_write(struct i40iw_qp_uk *qp,
op_info = &info->op.inline_rdma_write;
if (op_info->len > I40IW_MAX_INLINE_DATA_SIZE)
- return I40IW_ERR_INVALID_IMM_DATA_SIZE;
+ return I40IW_ERR_INVALID_INLINE_DATA_SIZE;
ret_code = i40iw_inline_data_size_to_wqesize(op_info->len, &wqe_size);
if (ret_code)
@@ -511,7 +511,7 @@ static enum i40iw_status_code i40iw_inline_send(struct i40iw_qp_uk *qp,
op_info = &info->op.inline_send;
if (op_info->len > I40IW_MAX_INLINE_DATA_SIZE)
- return I40IW_ERR_INVALID_IMM_DATA_SIZE;
+ return I40IW_ERR_INVALID_INLINE_DATA_SIZE;
ret_code = i40iw_inline_data_size_to_wqesize(op_info->len, &wqe_size);
if (ret_code)
@@ -1187,7 +1187,7 @@ enum i40iw_status_code i40iw_inline_data_size_to_wqesize(u32 data_size,
u8 *wqe_size)
{
if (data_size > I40IW_MAX_INLINE_DATA_SIZE)
- return I40IW_ERR_INVALID_IMM_DATA_SIZE;
+ return I40IW_ERR_INVALID_INLINE_DATA_SIZE;
if (data_size <= 16)
*wqe_size = I40IW_QP_WQE_MIN_SIZE;
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v2 3/6] i40iw: Improve CQP timeout logic
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2017-08-09 1:38 ` [PATCH v2 1/6] i40iw: Fix parsing of query/commit FPM buffers Henry Orosco
2017-08-09 1:38 ` [PATCH v2 2/6] i40iw: Correct variable names Henry Orosco
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-09 1:38 ` [PATCH v2 4/6] i40iw: Fix typecast of tcp_seq_num Henry Orosco
` (3 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f, Shiraz Saleem,
Henry Orosco
From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
The current timeout logic for Control Queue-Pair (CQP) OPs
does not take into account whether CQP makes progress but
rather blindly waits for a large timeout value, 100000 jiffies
for the completion event. Improve this by setting the timeout
based on whether the CQP is making progress or not. If the CQP
is hung, the timeout will happen sooner, in 5000 jiffies. Each
time the CQP progress is detetcted, the timeout extends by 5000
jiffies.
Signed-off-by: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Christopher N Bednarz <christopher.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 11 +++++++++++
drivers/infiniband/hw/i40iw/i40iw_p.h | 14 +++++++++-----
drivers/infiniband/hw/i40iw/i40iw_type.h | 5 +++++
drivers/infiniband/hw/i40iw/i40iw_utils.c | 22 ++++++++++++++--------
4 files changed, 39 insertions(+), 13 deletions(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index ef4a73c..99d9613 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -54,6 +54,17 @@ static inline void i40iw_insert_wqe_hdr(u64 *wqe, u64 header)
set_64bit_val(wqe, 24, header);
}
+void i40iw_check_cqp_progress(struct i40iw_cqp_timeout *cqp_timeout, struct i40iw_sc_dev *dev)
+{
+ if (cqp_timeout->compl_cqp_cmds != dev->cqp_cmd_stats[OP_COMPLETED_COMMANDS]) {
+ cqp_timeout->compl_cqp_cmds = dev->cqp_cmd_stats[OP_COMPLETED_COMMANDS];
+ cqp_timeout->count = 0;
+ } else {
+ if (dev->cqp_cmd_stats[OP_REQUESTED_COMMANDS] != cqp_timeout->compl_cqp_cmds)
+ cqp_timeout->count++;
+ }
+}
+
/**
* i40iw_get_cqp_reg_info - get head and tail for cqp using registers
* @cqp: struct for cqp hw
diff --git a/drivers/infiniband/hw/i40iw/i40iw_p.h b/drivers/infiniband/hw/i40iw/i40iw_p.h
index 28a92fe..e217a12 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_p.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_p.h
@@ -35,11 +35,13 @@
#ifndef I40IW_P_H
#define I40IW_P_H
-#define PAUSE_TIMER_VALUE 0xFFFF
-#define REFRESH_THRESHOLD 0x7FFF
-#define HIGH_THRESHOLD 0x800
-#define LOW_THRESHOLD 0x200
-#define ALL_TC2PFC 0xFF
+#define PAUSE_TIMER_VALUE 0xFFFF
+#define REFRESH_THRESHOLD 0x7FFF
+#define HIGH_THRESHOLD 0x800
+#define LOW_THRESHOLD 0x200
+#define ALL_TC2PFC 0xFF
+#define CQP_COMPL_WAIT_TIME 0x3E8
+#define CQP_TIMEOUT_THRESHOLD 5
void i40iw_debug_buf(struct i40iw_sc_dev *dev, enum i40iw_debug_flag mask,
char *desc, u64 *buf, u32 size);
@@ -51,6 +53,8 @@ void i40iw_sc_cqp_post_sq(struct i40iw_sc_cqp *cqp);
u64 *i40iw_sc_cqp_get_next_send_wqe(struct i40iw_sc_cqp *cqp, u64 scratch);
+void i40iw_check_cqp_progress(struct i40iw_cqp_timeout *cqp_timeout, struct i40iw_sc_dev *dev);
+
enum i40iw_status_code i40iw_sc_mr_fast_register(struct i40iw_sc_qp *qp,
struct i40iw_fast_reg_stag_info *info,
bool post_sq);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h
index 959ec81..63118f6 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_type.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_type.h
@@ -1345,4 +1345,9 @@ struct i40iw_virtchnl_work_info {
void *worker_vf_dev;
};
+struct i40iw_cqp_timeout {
+ u64 compl_cqp_cmds;
+ u8 count;
+};
+
#endif
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index e311ec5..62f1f45 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -445,23 +445,29 @@ static int i40iw_wait_event(struct i40iw_device *iwdev,
{
struct cqp_commands_info *info = &cqp_request->info;
struct i40iw_cqp *iwcqp = &iwdev->cqp;
+ struct i40iw_cqp_timeout cqp_timeout;
bool cqp_error = false;
int err_code = 0;
- int timeout_ret = 0;
+ memset(&cqp_timeout, 0, sizeof(cqp_timeout));
+ cqp_timeout.compl_cqp_cmds = iwdev->sc_dev.cqp_cmd_stats[OP_COMPLETED_COMMANDS];
+ do {
+ if (wait_event_timeout(cqp_request->waitq,
+ cqp_request->request_done, CQP_COMPL_WAIT_TIME))
+ break;
- timeout_ret = wait_event_timeout(cqp_request->waitq,
- cqp_request->request_done,
- I40IW_EVENT_TIMEOUT);
- if (!timeout_ret) {
- i40iw_pr_err("error cqp command 0x%x timed out ret = %d\n",
- info->cqp_cmd, timeout_ret);
+ i40iw_check_cqp_progress(&cqp_timeout, &iwdev->sc_dev);
+
+ if (cqp_timeout.count < CQP_TIMEOUT_THRESHOLD)
+ continue;
+
+ i40iw_pr_err("error cqp command 0x%x timed out", info->cqp_cmd);
err_code = -ETIME;
if (!iwdev->reset) {
iwdev->reset = true;
i40iw_request_reset(iwdev);
}
goto done;
- }
+ } while (1);
cqp_error = cqp_request->compl_info.error;
if (cqp_error) {
i40iw_pr_err("error cqp command 0x%x completion maj = 0x%x min=0x%x\n",
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v2 4/6] i40iw: Fix typecast of tcp_seq_num
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
` (2 preceding siblings ...)
2017-08-09 1:38 ` [PATCH v2 3/6] i40iw: Improve CQP timeout logic Henry Orosco
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-09 1:38 ` [PATCH v2 5/6] i40iw: Use correct alignment for CQ0 memory Henry Orosco
` (2 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f, Mustafa Ismail,
Henry Orosco
From: Mustafa Ismail <mustafa.ismail-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
The typecast of tcp_seq_num incorrectly uses u8. Fix by
casting to u32.
Signed-off-by: Mustafa Ismail <mustafa.ismail-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_uk.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 70a6b41..1060725 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -784,7 +784,7 @@ static enum i40iw_status_code i40iw_cq_poll_completion(struct i40iw_cq_uk *cq,
get_64bit_val(cqe, 0, &qword0);
get_64bit_val(cqe, 16, &qword2);
- info->tcp_seq_num = (u8)RS_64(qword0, I40IWCQ_TCPSEQNUM);
+ info->tcp_seq_num = (u32)RS_64(qword0, I40IWCQ_TCPSEQNUM);
info->qp_id = (u32)RS_64(qword2, I40IWCQ_QPID);
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v2 5/6] i40iw: Use correct alignment for CQ0 memory
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
` (3 preceding siblings ...)
2017-08-09 1:38 ` [PATCH v2 4/6] i40iw: Fix typecast of tcp_seq_num Henry Orosco
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-09 1:38 ` [PATCH v2 6/6] i40iw: Fix potential fcn_id_array out of bounds Henry Orosco
2017-08-16 15:29 ` [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw Doug Ledford
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f,
Christopher N Bednarz, Henry Orosco
From: Christopher N Bednarz <christopher.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Utilize correct alignment variable when allocating
DMA memory for CQ0.
Signed-off-by: Christopher N Bednarz <christopher.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_puda.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c
index 71050c5..7f5583d 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_puda.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c
@@ -685,7 +685,7 @@ static enum i40iw_status_code i40iw_puda_cq_create(struct i40iw_puda_rsrc *rsrc)
cqsize = rsrc->cq_size * (sizeof(struct i40iw_cqe));
tsize = cqsize + sizeof(struct i40iw_cq_shadow_area);
ret = i40iw_allocate_dma_mem(dev->hw, &rsrc->cqmem, tsize,
- I40IW_CQ0_ALIGNMENT_MASK);
+ I40IW_CQ0_ALIGNMENT);
if (ret)
return ret;
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v2 6/6] i40iw: Fix potential fcn_id_array out of bounds
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
` (4 preceding siblings ...)
2017-08-09 1:38 ` [PATCH v2 5/6] i40iw: Use correct alignment for CQ0 memory Henry Orosco
@ 2017-08-09 1:38 ` Henry Orosco
2017-08-16 15:29 ` [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw Doug Ledford
6 siblings, 0 replies; 9+ messages in thread
From: Henry Orosco @ 2017-08-09 1:38 UTC (permalink / raw)
To: dledford-H+wXaHxf7aLQT0dZR+AlfA
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f,
Christopher N Bednarz, Henry Orosco
From: Christopher N Bednarz <christopher.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Avoid out of bounds error by utilizing I40IW_MAX_STATS_COUNT
instead of I40IW_INVALID_FCN_ID.
Signed-off-by: Christopher N Bednarz <christoper.n.bednarz-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Henry Orosco <henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 99d9613..d1f5345 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -4892,7 +4892,7 @@ void i40iw_vsi_stats_free(struct i40iw_sc_vsi *vsi)
{
u8 fcn_id = vsi->fcn_id;
- if ((vsi->stats_fcn_id_alloc) && (fcn_id != I40IW_INVALID_FCN_ID))
+ if (vsi->stats_fcn_id_alloc && fcn_id < I40IW_MAX_STATS_COUNT)
vsi->dev->fcn_id_array[fcn_id] = false;
i40iw_hw_stats_stop_timer(vsi);
}
--
2.8.3
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
` (5 preceding siblings ...)
2017-08-09 1:38 ` [PATCH v2 6/6] i40iw: Fix potential fcn_id_array out of bounds Henry Orosco
@ 2017-08-16 15:29 ` Doug Ledford
[not found] ` <1502897345.33760.9.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
6 siblings, 1 reply; 9+ messages in thread
From: Doug Ledford @ 2017-08-16 15:29 UTC (permalink / raw)
To: Henry Orosco
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
On Tue, 2017-08-08 at 20:38 -0500, Henry Orosco wrote:
> The following patch series has fixes for the i40iw driver.
>
> V2 - Removed 1 patch from series
>
> --------------------------------------------------------
>
> Git Repo to pull from:
> https://github.com/horosco/linux.git
>
> tag:
> for-v4.13-08-07-2017-v2
>
> Commits of interest:
>
> commit 08bc3c4ee4c759655b2250b59ddec80950772f6a
> Date: Wed Jul 5 13:33:54 2017 -0500
> i40iw: Fix potential fcn_id_array out of bounds
>
> to
>
> commit a18153f9af05c1f03ee449cc881fd63e9f83d7fc
> Date: Mon Jul 3 15:45:32 2017 -0500
> i40iw: Fix parsing of query/commit FPM buffers
>
> --------------------------------------------------------
>
> Chien Tin Tung (1):
> i40iw: Fix parsing of query/commit FPM buffers
>
> Christopher N Bednarz (2):
> i40iw: Use correct alignment for CQ0 memory
> i40iw: Fix potential fcn_id_array out of bounds
>
> Mustafa Ismail (2):
> i40iw: Correct variable names
> i40iw: Fix typecast of tcp_seq_num
>
> Shiraz Saleem (1):
> i40iw: Improve CQP timeout logic
Hi Henry,
I took everything for -rc *Except* the Improve CQP timeout logic patch.
That patch qualifies as new development and not really a bug fix, and
since we are at -rc5, new development needs to be in for-next instead.
>
> drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 134
> ++++++++++++++++++++---------
> drivers/infiniband/hw/i40iw/i40iw_d.h | 4 +-
> drivers/infiniband/hw/i40iw/i40iw_p.h | 14 +--
> drivers/infiniband/hw/i40iw/i40iw_puda.c | 2 +-
> drivers/infiniband/hw/i40iw/i40iw_status.h | 2 +-
> drivers/infiniband/hw/i40iw/i40iw_type.h | 5 ++
> drivers/infiniband/hw/i40iw/i40iw_uk.c | 8 +-
> drivers/infiniband/hw/i40iw/i40iw_utils.c | 22 +++--
> 8 files changed, 129 insertions(+), 62 deletions(-)
>
--
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
GPG KeyID: B826A3330E572FDD
Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw
[not found] ` <1502897345.33760.9.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
@ 2017-08-22 19:39 ` Doug Ledford
0 siblings, 0 replies; 9+ messages in thread
From: Doug Ledford @ 2017-08-22 19:39 UTC (permalink / raw)
To: Henry Orosco
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA,
e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
On Wed, 2017-08-16 at 11:29 -0400, Doug Ledford wrote:
> Hi Henry,
>
> I took everything for -rc *Except* the Improve CQP timeout logic
> patch.
> That patch qualifies as new development and not really a bug fix,
> and
> since we are at -rc5, new development needs to be in for-next
> instead.
Hi Henry,
I've pulled the CQP timeout patch into for-next. Thanks.
--
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
GPG KeyID: B826A3330E572FDD
Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2017-08-22 19:39 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-08-09 1:38 [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw Henry Orosco
[not found] ` <20170809013848.29080-1-henry.orosco-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2017-08-09 1:38 ` [PATCH v2 1/6] i40iw: Fix parsing of query/commit FPM buffers Henry Orosco
2017-08-09 1:38 ` [PATCH v2 2/6] i40iw: Correct variable names Henry Orosco
2017-08-09 1:38 ` [PATCH v2 3/6] i40iw: Improve CQP timeout logic Henry Orosco
2017-08-09 1:38 ` [PATCH v2 4/6] i40iw: Fix typecast of tcp_seq_num Henry Orosco
2017-08-09 1:38 ` [PATCH v2 5/6] i40iw: Use correct alignment for CQ0 memory Henry Orosco
2017-08-09 1:38 ` [PATCH v2 6/6] i40iw: Fix potential fcn_id_array out of bounds Henry Orosco
2017-08-16 15:29 ` [pull request][PATCH v2 0/6] i40iw: Fixes for i40iw Doug Ledford
[not found] ` <1502897345.33760.9.camel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2017-08-22 19:39 ` Doug Ledford
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.