All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
	Tj <ml.linux@elloe.vision>, Joerg Roedel <jroedel@suse.de>,
	Sasha Levin <sashal@kernel.org>,
	iommu@lists.linux-foundation.org
Subject: [PATCH AUTOSEL 5.10 21/47] iommu/amd: Fix performance counter initialization
Date: Tue,  2 Mar 2021 06:56:20 -0500	[thread overview]
Message-ID: <20210302115646.62291-21-sashal@kernel.org> (raw)
In-Reply-To: <20210302115646.62291-1-sashal@kernel.org>

From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>

[ Upstream commit 6778ff5b21bd8e78c8bd547fd66437cf2657fd9b ]

Certain AMD platforms enable power gating feature for IOMMU PMC,
which prevents the IOMMU driver from updating the counter while
trying to validate the PMC functionality in the init_iommu_perf_ctr().
This results in disabling PMC support and the following error message:

    "AMD-Vi: Unable to read/write to IOMMU perf counter"

To workaround this issue, disable power gating temporarily by programming
the counter source to non-zero value while validating the counter,
and restore the prior state afterward.

Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Tested-by: Tj (Elloe Linux) <ml.linux@elloe.vision>
Link: https://lore.kernel.org/r/20210208122712.5048-1-suravee.suthikulpanit@amd.com
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=201753
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/iommu/amd/init.c | 45 ++++++++++++++++++++++++++++++----------
 1 file changed, 34 insertions(+), 11 deletions(-)

diff --git a/drivers/iommu/amd/init.c b/drivers/iommu/amd/init.c
index c842545368fd..3c215f0a6052 100644
--- a/drivers/iommu/amd/init.c
+++ b/drivers/iommu/amd/init.c
@@ -12,6 +12,7 @@
 #include <linux/acpi.h>
 #include <linux/list.h>
 #include <linux/bitmap.h>
+#include <linux/delay.h>
 #include <linux/slab.h>
 #include <linux/syscore_ops.h>
 #include <linux/interrupt.h>
@@ -254,6 +255,8 @@ static enum iommu_init_state init_state = IOMMU_START_STATE;
 static int amd_iommu_enable_interrupts(void);
 static int __init iommu_go_to_state(enum iommu_init_state state);
 static void init_device_table_dma(void);
+static int iommu_pc_get_set_reg(struct amd_iommu *iommu, u8 bank, u8 cntr,
+				u8 fxn, u64 *value, bool is_write);
 
 static bool amd_iommu_pre_enabled = true;
 
@@ -1717,13 +1720,11 @@ static int __init init_iommu_all(struct acpi_table_header *table)
 	return 0;
 }
 
-static int iommu_pc_get_set_reg(struct amd_iommu *iommu, u8 bank, u8 cntr,
-				u8 fxn, u64 *value, bool is_write);
-
-static void init_iommu_perf_ctr(struct amd_iommu *iommu)
+static void __init init_iommu_perf_ctr(struct amd_iommu *iommu)
 {
+	int retry;
 	struct pci_dev *pdev = iommu->dev;
-	u64 val = 0xabcd, val2 = 0, save_reg = 0;
+	u64 val = 0xabcd, val2 = 0, save_reg, save_src;
 
 	if (!iommu_feature(iommu, FEATURE_PC))
 		return;
@@ -1731,17 +1732,39 @@ static void init_iommu_perf_ctr(struct amd_iommu *iommu)
 	amd_iommu_pc_present = true;
 
 	/* save the value to restore, if writable */
-	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, false))
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, false) ||
+	    iommu_pc_get_set_reg(iommu, 0, 0, 8, &save_src, false))
 		goto pc_false;
 
-	/* Check if the performance counters can be written to */
-	if ((iommu_pc_get_set_reg(iommu, 0, 0, 0, &val, true)) ||
-	    (iommu_pc_get_set_reg(iommu, 0, 0, 0, &val2, false)) ||
-	    (val != val2))
+	/*
+	 * Disable power gating by programing the performance counter
+	 * source to 20 (i.e. counts the reads and writes from/to IOMMU
+	 * Reserved Register [MMIO Offset 1FF8h] that are ignored.),
+	 * which never get incremented during this init phase.
+	 * (Note: The event is also deprecated.)
+	 */
+	val = 20;
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 8, &val, true))
 		goto pc_false;
 
+	/* Check if the performance counters can be written to */
+	val = 0xabcd;
+	for (retry = 5; retry; retry--) {
+		if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &val, true) ||
+		    iommu_pc_get_set_reg(iommu, 0, 0, 0, &val2, false) ||
+		    val2)
+			break;
+
+		/* Wait about 20 msec for power gating to disable and retry. */
+		msleep(20);
+	}
+
 	/* restore */
-	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, true))
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, true) ||
+	    iommu_pc_get_set_reg(iommu, 0, 0, 8, &save_src, true))
+		goto pc_false;
+
+	if (val != val2)
 		goto pc_false;
 
 	pci_info(pdev, "IOMMU performance counters supported\n");
-- 
2.30.1


WARNING: multiple messages have this Message-ID (diff)
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Sasha Levin <sashal@kernel.org>, Tj <ml.linux@elloe.vision>,
	iommu@lists.linux-foundation.org, Joerg Roedel <jroedel@suse.de>
Subject: [PATCH AUTOSEL 5.10 21/47] iommu/amd: Fix performance counter initialization
Date: Tue,  2 Mar 2021 06:56:20 -0500	[thread overview]
Message-ID: <20210302115646.62291-21-sashal@kernel.org> (raw)
In-Reply-To: <20210302115646.62291-1-sashal@kernel.org>

From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>

[ Upstream commit 6778ff5b21bd8e78c8bd547fd66437cf2657fd9b ]

Certain AMD platforms enable power gating feature for IOMMU PMC,
which prevents the IOMMU driver from updating the counter while
trying to validate the PMC functionality in the init_iommu_perf_ctr().
This results in disabling PMC support and the following error message:

    "AMD-Vi: Unable to read/write to IOMMU perf counter"

To workaround this issue, disable power gating temporarily by programming
the counter source to non-zero value while validating the counter,
and restore the prior state afterward.

Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Tested-by: Tj (Elloe Linux) <ml.linux@elloe.vision>
Link: https://lore.kernel.org/r/20210208122712.5048-1-suravee.suthikulpanit@amd.com
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=201753
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/iommu/amd/init.c | 45 ++++++++++++++++++++++++++++++----------
 1 file changed, 34 insertions(+), 11 deletions(-)

diff --git a/drivers/iommu/amd/init.c b/drivers/iommu/amd/init.c
index c842545368fd..3c215f0a6052 100644
--- a/drivers/iommu/amd/init.c
+++ b/drivers/iommu/amd/init.c
@@ -12,6 +12,7 @@
 #include <linux/acpi.h>
 #include <linux/list.h>
 #include <linux/bitmap.h>
+#include <linux/delay.h>
 #include <linux/slab.h>
 #include <linux/syscore_ops.h>
 #include <linux/interrupt.h>
@@ -254,6 +255,8 @@ static enum iommu_init_state init_state = IOMMU_START_STATE;
 static int amd_iommu_enable_interrupts(void);
 static int __init iommu_go_to_state(enum iommu_init_state state);
 static void init_device_table_dma(void);
+static int iommu_pc_get_set_reg(struct amd_iommu *iommu, u8 bank, u8 cntr,
+				u8 fxn, u64 *value, bool is_write);
 
 static bool amd_iommu_pre_enabled = true;
 
@@ -1717,13 +1720,11 @@ static int __init init_iommu_all(struct acpi_table_header *table)
 	return 0;
 }
 
-static int iommu_pc_get_set_reg(struct amd_iommu *iommu, u8 bank, u8 cntr,
-				u8 fxn, u64 *value, bool is_write);
-
-static void init_iommu_perf_ctr(struct amd_iommu *iommu)
+static void __init init_iommu_perf_ctr(struct amd_iommu *iommu)
 {
+	int retry;
 	struct pci_dev *pdev = iommu->dev;
-	u64 val = 0xabcd, val2 = 0, save_reg = 0;
+	u64 val = 0xabcd, val2 = 0, save_reg, save_src;
 
 	if (!iommu_feature(iommu, FEATURE_PC))
 		return;
@@ -1731,17 +1732,39 @@ static void init_iommu_perf_ctr(struct amd_iommu *iommu)
 	amd_iommu_pc_present = true;
 
 	/* save the value to restore, if writable */
-	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, false))
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, false) ||
+	    iommu_pc_get_set_reg(iommu, 0, 0, 8, &save_src, false))
 		goto pc_false;
 
-	/* Check if the performance counters can be written to */
-	if ((iommu_pc_get_set_reg(iommu, 0, 0, 0, &val, true)) ||
-	    (iommu_pc_get_set_reg(iommu, 0, 0, 0, &val2, false)) ||
-	    (val != val2))
+	/*
+	 * Disable power gating by programing the performance counter
+	 * source to 20 (i.e. counts the reads and writes from/to IOMMU
+	 * Reserved Register [MMIO Offset 1FF8h] that are ignored.),
+	 * which never get incremented during this init phase.
+	 * (Note: The event is also deprecated.)
+	 */
+	val = 20;
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 8, &val, true))
 		goto pc_false;
 
+	/* Check if the performance counters can be written to */
+	val = 0xabcd;
+	for (retry = 5; retry; retry--) {
+		if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &val, true) ||
+		    iommu_pc_get_set_reg(iommu, 0, 0, 0, &val2, false) ||
+		    val2)
+			break;
+
+		/* Wait about 20 msec for power gating to disable and retry. */
+		msleep(20);
+	}
+
 	/* restore */
-	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, true))
+	if (iommu_pc_get_set_reg(iommu, 0, 0, 0, &save_reg, true) ||
+	    iommu_pc_get_set_reg(iommu, 0, 0, 8, &save_src, true))
+		goto pc_false;
+
+	if (val != val2)
 		goto pc_false;
 
 	pci_info(pdev, "IOMMU performance counters supported\n");
-- 
2.30.1

_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu

  parent reply	other threads:[~2021-03-02 19:48 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-02 11:56 [PATCH AUTOSEL 5.10 01/47] i2c: rcar: faster irq code to minimize HW race condition Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 02/47] i2c: rcar: optimize cacheline " Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 03/47] scsi: ufs: Add a quirk to permit overriding UniPro defaults Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 04/47] scsi: ufs: WB is only available on LUN #0 to #7 Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 05/47] scsi: ufs: Introduce a quirk to allow only page-aligned sg entries Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 06/47] udf: fix silent AED tagLocation corruption Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 07/47] iommu/vt-d: Clear PRQ overflow only when PRQ is empty Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 08/47] mmc: mxs-mmc: Fix a resource leak in an error handling path in 'mxs_mmc_probe()' Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 09/47] mmc: sdhci-of-dwcmshc: set SDHCI_QUIRK2_PRESET_VALUE_BROKEN Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 10/47] mmc: mediatek: fix race condition between msdc_request_timeout and irq Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 11/47] mmc: sdhci-iproc: Add ACPI bindings for the RPi Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 16:16   ` Jeremy Linton
2021-03-02 16:16     ` Jeremy Linton
2021-03-12 22:12     ` Sasha Levin
2021-03-12 22:12       ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 12/47] Platform: OLPC: Fix probe error handling Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 13/47] powerpc/pci: Add ppc_md.discover_phbs() Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 14/47] HID: i2c-hid: Add I2C_HID_QUIRK_NO_IRQ_AFTER_RESET for ITE8568 EC on Voyo Winpad A15 Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 15/47] spi: stm32: make spurious and overrun interrupts visible Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 16/47] powerpc: improve handling of unrecoverable system reset Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 17/47] powerpc/perf: Record counter overflow always if SAMPLE_IP is unset Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 18/47] kunit: tool: fix unit test cleanup handling Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 19/47] HID: logitech-dj: add support for the new lightspeed connection iteration Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 20/47] powerpc/64: Fix stack trace not displaying final frame Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` Sasha Levin [this message]
2021-03-02 11:56   ` [PATCH AUTOSEL 5.10 21/47] iommu/amd: Fix performance counter initialization Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 22/47] clk: qcom: gdsc: Implement NO_RET_PERIPH flag Sasha Levin
2021-03-02 23:04   ` Stephen Boyd
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 23/47] sparc32: Limit memblock allocation to low memory Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 24/47] sparc64: Use arch_validate_flags() to validate ADI flag Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 25/47] ACPICA: Fix race in generic_serial_bus (I2C) and GPIO op_region parameter handling Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 26/47] Input: applespi - don't wait for responses to commands indefinitely Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 27/47] x86, build: use objtool mcount Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 28/47] PCI: xgene-msi: Fix race in installing chained irq handler Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 29/47] misc: eeprom_93xx46: Add quirk to support Microchip 93LC46B eeprom Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 30/47] PCI: mediatek: Add missing of_node_put() to fix reference leak Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 31/47] drivers/base: build kunit tests without structleak plugin Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 32/47] drm/msm/a5xx: Remove overwriting A5XX_PC_DBG_ECO_CNTL register Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 33/47] PCI/LINK: Remove bandwidth notification Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 34/47] ext4: don't try to processed freed blocks until mballoc is initialized Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 35/47] PCI: cadence: Retrain Link to work around Gen2 training defect Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 36/47] kbuild: clamp SUBLEVEL to 255 Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 37/47] PCI: Fix pci_register_io_range() memory leak Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 38/47] i40e: Fix memory leak in i40e_probe Sasha Levin
2021-03-02 11:56   ` [Intel-wired-lan] " Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 39/47] kasan: fix memory corruption in kasan_bitops_tags test Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 40/47] riscv: Get rid of MAX_EARLY_MAPPING_SIZE Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 41/47] s390/smp: __smp_rescan_cpus() - move cpumask away from stack Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 42/47] drivers/base/memory: don't store phys_device in memory blocks Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 43/47] sysctl.c: fix underflow value setting risk in vm_table Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 44/47] scsi: libiscsi: Fix iscsi_prep_scsi_cmd_pdu() error handling Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 45/47] nbd: handle device refs for DESTROY_ON_DISCONNECT properly Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 46/47] scsi: target: core: Add cmd length set before cmd complete Sasha Levin
2021-03-02 11:56   ` Sasha Levin
2021-03-02 11:56 ` [PATCH AUTOSEL 5.10 47/47] scsi: target: core: Prevent underflow for service actions 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=20210302115646.62291-21-sashal@kernel.org \
    --to=sashal@kernel.org \
    --cc=iommu@lists.linux-foundation.org \
    --cc=jroedel@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=ml.linux@elloe.vision \
    --cc=stable@vger.kernel.org \
    --cc=suravee.suthikulpanit@amd.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 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.