linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Vijay Immanuel <vijayi@attalasystems.com>,
	Doug Ledford <dledford@redhat.com>,
	Sasha Levin <sashal@kernel.org>,
	linux-rdma@vger.kernel.org
Subject: [PATCH AUTOSEL 4.9 17/64] IB/rxe: fixes for rdma read retry
Date: Fri,  8 Nov 2019 06:44:58 -0500	[thread overview]
Message-ID: <20191108114545.15351-17-sashal@kernel.org> (raw)
In-Reply-To: <20191108114545.15351-1-sashal@kernel.org>

From: Vijay Immanuel <vijayi@attalasystems.com>

[ Upstream commit 030e46e495af855a13964a0aab9753ea82a96edc ]

When a read request is retried for the remaining partial
data, the response may restart from read response first
or read response only. So support those cases.

Do not advance the comp psn beyond the current wqe's last_psn
as that could skip over an entire read wqe and will cause the
req_retry() logic to set an incorrect req psn.
An example sequence is as follows:
Write        PSN 40 -- this is the current WQE.
Read request PSN 41
Write        PSN 42
Receive ACK  PSN 42 -- this will complete the current WQE
for PSN 40, and set the comp psn to 42 which is a problem
because the read request at PSN 41 has been skipped over.
So when req_retry() tries to retransmit the read request,
it sets the req psn to 42 which is incorrect.

When retrying a read request, calculate the number of psns
completed based on the dma resid instead of the wqe first_psn.
The wqe first_psn could have moved if the read request was
retried multiple times.

Set the reth length to the dma resid to handle read retries for
the remaining partial data.

Signed-off-by: Vijay Immanuel <vijayi@attalasystems.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/infiniband/sw/rxe/rxe_comp.c | 21 ++++++++++++++++-----
 drivers/infiniband/sw/rxe/rxe_req.c  | 15 +++++++++------
 2 files changed, 25 insertions(+), 11 deletions(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
index df15b6d7b645e..b03a6206d9be7 100644
--- a/drivers/infiniband/sw/rxe/rxe_comp.c
+++ b/drivers/infiniband/sw/rxe/rxe_comp.c
@@ -250,6 +250,17 @@ static inline enum comp_state check_ack(struct rxe_qp *qp,
 	case IB_OPCODE_RC_RDMA_READ_RESPONSE_MIDDLE:
 		if (pkt->opcode != IB_OPCODE_RC_RDMA_READ_RESPONSE_MIDDLE &&
 		    pkt->opcode != IB_OPCODE_RC_RDMA_READ_RESPONSE_LAST) {
+			/* read retries of partial data may restart from
+			 * read response first or response only.
+			 */
+			if ((pkt->psn == wqe->first_psn &&
+			     pkt->opcode ==
+			     IB_OPCODE_RC_RDMA_READ_RESPONSE_FIRST) ||
+			    (wqe->first_psn == wqe->last_psn &&
+			     pkt->opcode ==
+			     IB_OPCODE_RC_RDMA_READ_RESPONSE_ONLY))
+				break;
+
 			return COMPST_ERROR;
 		}
 		break;
@@ -486,11 +497,11 @@ static inline enum comp_state complete_wqe(struct rxe_qp *qp,
 					   struct rxe_pkt_info *pkt,
 					   struct rxe_send_wqe *wqe)
 {
-	qp->comp.opcode = -1;
-
-	if (pkt) {
-		if (psn_compare(pkt->psn, qp->comp.psn) >= 0)
-			qp->comp.psn = (pkt->psn + 1) & BTH_PSN_MASK;
+	if (pkt && wqe->state == wqe_state_pending) {
+		if (psn_compare(wqe->last_psn, qp->comp.psn) >= 0) {
+			qp->comp.psn = (wqe->last_psn + 1) & BTH_PSN_MASK;
+			qp->comp.opcode = -1;
+		}
 
 		if (qp->req.wait_psn) {
 			qp->req.wait_psn = 0;
diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 0f9fe2ca2a91a..6fb771290c566 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -72,9 +72,6 @@ static void req_retry(struct rxe_qp *qp)
 	int npsn;
 	int first = 1;
 
-	wqe = queue_head(qp->sq.queue);
-	npsn = (qp->comp.psn - wqe->first_psn) & BTH_PSN_MASK;
-
 	qp->req.wqe_index	= consumer_index(qp->sq.queue);
 	qp->req.psn		= qp->comp.psn;
 	qp->req.opcode		= -1;
@@ -106,11 +103,17 @@ static void req_retry(struct rxe_qp *qp)
 		if (first) {
 			first = 0;
 
-			if (mask & WR_WRITE_OR_SEND_MASK)
+			if (mask & WR_WRITE_OR_SEND_MASK) {
+				npsn = (qp->comp.psn - wqe->first_psn) &
+					BTH_PSN_MASK;
 				retry_first_write_send(qp, wqe, mask, npsn);
+			}
 
-			if (mask & WR_READ_MASK)
+			if (mask & WR_READ_MASK) {
+				npsn = (wqe->dma.length - wqe->dma.resid) /
+					qp->mtu;
 				wqe->iova += npsn * qp->mtu;
+			}
 		}
 
 		wqe->state = wqe_state_posted;
@@ -439,7 +442,7 @@ static struct sk_buff *init_req_packet(struct rxe_qp *qp,
 	if (pkt->mask & RXE_RETH_MASK) {
 		reth_set_rkey(pkt, ibwr->wr.rdma.rkey);
 		reth_set_va(pkt, wqe->iova);
-		reth_set_len(pkt, wqe->dma.length);
+		reth_set_len(pkt, wqe->dma.resid);
 	}
 
 	if (pkt->mask & RXE_IMMDT_MASK)
-- 
2.20.1


  parent reply	other threads:[~2019-11-08 11:54 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-08 11:44 [PATCH AUTOSEL 4.9 01/64] ath10k: fix kernel panic by moving pci flush after napi_disable Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 02/64] iio: dac: mcp4922: fix error handling in mcp4922_write_raw Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 03/64] ALSA: pcm: signedness bug in snd_pcm_plug_alloc() Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 04/64] arm64: dts: tegra210-p2180: Correct sdmmc4 vqmmc-supply Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 05/64] ARM: dts: at91/trivial: Fix USART1 definition for at91sam9g45 Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 06/64] cfg80211: Avoid regulatory restore when COUNTRY_IE_IGNORE is set Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 07/64] ALSA: seq: Do error checks at creating system ports Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 08/64] ath9k: fix tx99 with monitor mode interface Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 09/64] gfs2: Don't set GFS2_RDF_UPTODATE when the lvb is updated Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 10/64] ASoC: dpcm: Properly initialise hw->rate_max Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 11/64] MIPS: BCM47XX: Enable USB power on Netgear WNDR3400v3 Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 12/64] ARM: dts: exynos: Fix sound in Snow-rev5 Chromebook Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 13/64] ARM: dts: exynos: Fix regulators configuration on Peach Pi/Pit Chromebooks Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 14/64] i40e: use correct length for strncpy Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 15/64] i40e: hold the rtnl lock on clearing interrupt scheme Sasha Levin
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 16/64] i40e: Prevent deleting MAC address from VF when set by PF Sasha Levin
2019-11-08 11:44 ` Sasha Levin [this message]
2019-11-08 11:44 ` [PATCH AUTOSEL 4.9 18/64] iwlwifi: mvm: avoid sending too many BARs Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 19/64] ARM: dts: pxa: fix power i2c base address Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 20/64] rtl8187: Fix warning generated when strncpy() destination length matches the sixe argument Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 21/64] net: lan78xx: Bail out if lan78xx_get_endpoints fails Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 22/64] ASoC: sgtl5000: avoid division by zero if lo_vag is zero Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 23/64] ARM: dts: exynos: Disable pull control for S5M8767 PMIC Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 24/64] ath10k: wmi: disable softirq's while calling ieee80211_rx Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 25/64] mips: txx9: fix iounmap related issue Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 26/64] ASoC: Intel: hdac_hdmi: Limit sampling rates at dai creation Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 27/64] of: make PowerMac cache node search conditional on CONFIG_PPC_PMAC Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 28/64] ARM: dts: omap3-gta04: give spi_lcd node a label so that we can overwrite in other DTS files Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 29/64] ARM: dts: omap3-gta04: fixes for tvout / venc Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 30/64] ARM: dts: omap3-gta04: tvout: enable as display1 alias Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 31/64] ARM: dts: omap3-gta04: fix touchscreen tsc2007 Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 32/64] ARM: dts: omap3-gta04: make NAND partitions compatible with recent U-Boot Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 33/64] ARM: dts: omap3-gta04: keep vpll2 always on Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 34/64] dmaengine: dma-jz4780: Don't depend on MACH_JZ4780 Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 35/64] dmaengine: dma-jz4780: Further residue status fix Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 36/64] ath9k: add back support for using active monitor interfaces for tx99 Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 37/64] signal: Always ignore SIGKILL and SIGSTOP sent to the global init Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 38/64] signal: Properly deliver SIGILL from uprobes Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 39/64] signal: Properly deliver SIGSEGV from x86 uprobes Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 40/64] f2fs: fix memory leak of percpu counter in fill_super() Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 41/64] scsi: sym53c8xx: fix NULL pointer dereference panic in sym_int_sir() Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 42/64] ARM: imx6: register pm_power_off handler if "fsl,pmic-stby-poweroff" is set Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 43/64] scsi: pm80xx: Corrected dma_unmap_sg() parameter Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 44/64] scsi: pm80xx: Fixed system hang issue during kexec boot Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 45/64] kprobes: Don't call BUG_ON() if there is a kprobe in use on free list Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 46/64] nvmem: core: return error code instead of NULL from nvmem_device_get Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 47/64] media: fix: media: pci: meye: validate offset to avoid arbitrary access Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 48/64] media: dvb: fix compat ioctl translation Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 49/64] ALSA: intel8x0m: Register irq handler after register initializations Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 50/64] pinctrl: at91-pio4: fix has_config check in atmel_pctl_dt_subnode_to_map() Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 51/64] llc: avoid blocking in llc_sap_close() Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 52/64] ARM: dts: qcom: ipq4019: fix cpu0's qcom,saw2 reg value Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 53/64] powerpc/vdso: Correct call frame information Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 54/64] ARM: dts: socfpga: Fix I2C bus unit-address error Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 55/64] pinctrl: at91: don't use the same irqchip with multiple gpiochips Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 56/64] cxgb4: Fix endianness issue in t4_fwcache() Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 57/64] power: supply: ab8500_fg: silence uninitialized variable warnings Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 58/64] power: reset: at91-poweroff: do not procede if at91_shdwc is allocated Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 59/64] power: supply: max8998-charger: Fix platform data retrieval Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 60/64] component: fix loop condition to call unbind() if bind() fails Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 61/64] kernfs: Fix range checks in kernfs_get_target_path Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 62/64] ip_gre: fix parsing gre header in ipgre_err Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 63/64] ARM: dts: rockchip: Fix erroneous SPI bus dtc warnings on rk3036 Sasha Levin
2019-11-08 11:45 ` [PATCH AUTOSEL 4.9 64/64] ath9k: Fix a locking bug in ath9k_add_interface() Sasha Levin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20191108114545.15351-17-sashal@kernel.org \
    --to=sashal@kernel.org \
    --cc=dledford@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=vijayi@attalasystems.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).