* [dpdk-dev] [PATCH 1/4] [PATCH 1/4] igb_uio: Add interrupt_mode sysfs entry for igb_uio devices
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
@ 2014-06-05 14:39 ` Alan Carew
2014-06-05 17:20 ` Stephen Hemminger
2014-06-05 14:39 ` [dpdk-dev] [PATCH 2/4] [PATCH 2/4] eal_pci: Add interrupt mode to rte_pci_device and parsing to eal_pci Alan Carew
` (5 subsequent siblings)
6 siblings, 1 reply; 9+ messages in thread
From: Alan Carew @ 2014-06-05 14:39 UTC (permalink / raw)
To: dev
This patch adds an "interrupt_mode" sysfs entry for igb_uio devices,
allowing userspace eal_pci to track which interrupt mode has been enabled in kernel space.
The sysfs entry can be inspected via /sys/bus/pci/devices/<D:B:D:F/interrupt_mode
Signed-off-by: Alan Carew <alan.carew@intel.com>
---
.../common/include/rte_pci_dev_feature_defs.h | 85 ++++++++++++++++++++
.../common/include/rte_pci_dev_features.h | 70 ++++++++++++++++
lib/librte_eal/linuxapp/igb_uio/igb_uio.c | 48 +++++++++---
3 files changed, 193 insertions(+), 10 deletions(-)
create mode 100755 lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
create mode 100755 lib/librte_eal/common/include/rte_pci_dev_features.h
diff --git a/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h b/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
new file mode 100755
index 0000000..d23ed7d
--- /dev/null
+++ b/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
@@ -0,0 +1,85 @@
+/*-
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.GPL.
+ *
+ * Contact Information:
+ * Intel Corporation
+ *
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#ifndef _RTE_PCI_DEV_DEFS_H_
+#define _RTE_PCI_DEV_DEFS_H_
+
+#define RTE_PCI_DEV_FEATURE_INTR_MODE "interrupt_mode"
+
+#define INTR_NAME_LEN 10
+
+#define IGBUIO_NONE_INTR_NAME "none"
+#define IGBUIO_LEGACY_INTR_NAME "legacy"
+#define IGBUIO_MSI_INTR_NAME "msi"
+#define IGBUIO_MSIX_INTR_NAME "msix"
+
+
+#define INTR_MODE(id, mode_name) \
+ .mode = (id), .name = (mode_name)
+
+/* interrupt mode */
+enum igbuio_intr_mode {
+ IGBUIO_NONE_INTR_MODE = 0,
+ IGBUIO_LEGACY_INTR_MODE,
+ IGBUIO_MSI_INTR_MODE,
+ IGBUIO_MSIX_INTR_MODE,
+ IGBUIO_INTR_MODE_MAX
+};
+
+#endif /* _RTE_PCI_DEV_DEFS_H_ */
diff --git a/lib/librte_eal/common/include/rte_pci_dev_features.h b/lib/librte_eal/common/include/rte_pci_dev_features.h
new file mode 100755
index 0000000..a45c056
--- /dev/null
+++ b/lib/librte_eal/common/include/rte_pci_dev_features.h
@@ -0,0 +1,70 @@
+/*-
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ * The full GNU General Public License is included in this distribution
+ * in the file called LICENSE.GPL.
+ *
+ * Contact Information:
+ * Intel Corporation
+ *
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#ifndef RTE_PCI_DEV_INTR_MODE
+#define RTE_PCI_DEV_INTR_MODE(id, mode_name)
+#endif
+
+RTE_PCI_DEV_INTR_MODE(IGBUIO_NONE_INTR_MODE, IGBUIO_NONE_INTR_NAME)
+RTE_PCI_DEV_INTR_MODE(IGBUIO_LEGACY_INTR_MODE, IGBUIO_LEGACY_INTR_NAME)
+RTE_PCI_DEV_INTR_MODE(IGBUIO_MSI_INTR_MODE, IGBUIO_MSI_INTR_NAME)
+RTE_PCI_DEV_INTR_MODE(IGBUIO_MSIX_INTR_MODE, IGBUIO_MSIX_INTR_NAME)
+
+#undef RTE_PCI_DEV_INTR_MODE
+
diff --git a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
index 09c40bf..dd1124b 100644
--- a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
+++ b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
@@ -34,6 +34,8 @@
#include <xen/xen.h>
#endif
+#include <rte_pci_dev_feature_defs.h>
+
/**
* MSI-X related macros, copy from linux/pci_regs.h in kernel 2.6.39,
* but none of them in kernel 2.6.35.
@@ -49,12 +51,15 @@
#define IGBUIO_NUM_MSI_VECTORS 1
-/* interrupt mode */
-enum igbuio_intr_mode {
- IGBUIO_LEGACY_INTR_MODE = 0,
- IGBUIO_MSI_INTR_MODE,
- IGBUIO_MSIX_INTR_MODE,
- IGBUIO_INTR_MODE_MAX
+struct rte_pci_dev_intr_mode {
+ enum igbuio_intr_mode mode;
+ const char *name;
+};
+
+/* Table of interrupt modes */
+static struct rte_pci_dev_intr_mode interrupt_modes[] = {
+#define RTE_PCI_DEV_INTR_MODE(id, mode_name) {INTR_MODE(id, mode_name)},
+#include <rte_pci_dev_features.h>
};
/**
@@ -150,8 +155,32 @@ store_max_vfs(struct device *dev, struct device_attribute *attr,
}
static DEVICE_ATTR(max_vfs, S_IRUGO | S_IWUSR, show_max_vfs, store_max_vfs);
+
+static ssize_t
+show_interrupt_mode(struct device *dev, struct device_attribute *attr,
+ char *buf)
+{
+ unsigned i, num_modes;
+ enum igbuio_intr_mode mode;
+ struct uio_info *info;
+ struct pci_dev *pdev = container_of(dev, struct pci_dev, dev);
+
+ info = pci_get_drvdata(pdev);
+ mode = ((struct rte_uio_pci_dev *)info->priv)->mode;
+ num_modes = sizeof(interrupt_modes)/sizeof(interrupt_modes[0]);
+
+ for (i = 0; i < num_modes; i++) {
+ if (mode == interrupt_modes[i].mode)
+ return snprintf(buf, INTR_NAME_LEN, "%s\n", interrupt_modes[i].name );
+ }
+ return snprintf(buf, INTR_NAME_LEN, "unknown\n");
+}
+
+static DEVICE_ATTR(interrupt_mode, S_IRUGO, show_interrupt_mode, NULL);
+
static struct attribute *dev_attrs[] = {
&dev_attr_max_vfs.attr,
+ &dev_attr_interrupt_mode.attr,
NULL,
};
@@ -520,7 +549,7 @@ igbuio_pci_probe(struct pci_dev *dev, const struct pci_device_id *id)
#endif
udev->info.priv = udev;
udev->pdev = dev;
- udev->mode = 0; /* set the default value for interrupt mode */
+ udev->mode = IGBUIO_LEGACY_INTR_MODE; /* set the default value for interrupt mode */
spin_lock_init(&udev->lock);
/* check if it need to try msix first */
@@ -610,11 +639,10 @@ igbuio_config_intr_mode(char *intr_str)
printk(KERN_INFO "Use MSIX interrupt by default\n");
return 0;
}
-
- if (!strcmp(intr_str, "msix")) {
+ if (!strcmp(intr_str, IGBUIO_MSIX_INTR_NAME)) {
igbuio_intr_mode_preferred = IGBUIO_MSIX_INTR_MODE;
printk(KERN_INFO "Use MSIX interrupt\n");
- } else if (!strcmp(intr_str, "legacy")) {
+ } else if (!strcmp(intr_str, IGBUIO_LEGACY_INTR_NAME)) {
igbuio_intr_mode_preferred = IGBUIO_LEGACY_INTR_MODE;
printk(KERN_INFO "Use legacy interrupt\n");
} else {
--
1.7.0.7
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [dpdk-dev] [PATCH 1/4] [PATCH 1/4] igb_uio: Add interrupt_mode sysfs entry for igb_uio devices
2014-06-05 14:39 ` [dpdk-dev] [PATCH 1/4] [PATCH 1/4] igb_uio: Add interrupt_mode sysfs entry for igb_uio devices Alan Carew
@ 2014-06-05 17:20 ` Stephen Hemminger
0 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2014-06-05 17:20 UTC (permalink / raw)
To: Alan Carew; +Cc: dev
On Thu, 5 Jun 2014 15:39:16 +0100
Alan Carew <alan.carew@intel.com> wrote:
> This patch adds an "interrupt_mode" sysfs entry for igb_uio devices,
> allowing userspace eal_pci to track which interrupt mode has been enabled in kernel space.
> The sysfs entry can be inspected via /sys/bus/pci/devices/<D:B:D:F/interrupt_mode
>
> Signed-off-by: Alan Carew <alan.carew@intel.com>
> ---
> .../common/include/rte_pci_dev_feature_defs.h | 85 ++++++++++++++++++++
> .../common/include/rte_pci_dev_features.h | 70 ++++++++++++++++
> lib/librte_eal/linuxapp/igb_uio/igb_uio.c | 48 +++++++++---
> 3 files changed, 193 insertions(+), 10 deletions(-)
> create mode 100755 lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
> create mode 100755 lib/librte_eal/common/include/rte_pci_dev_features.h
>
> diff --git a/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h b/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
> new file mode 100755
> index 0000000..d23ed7d
> --- /dev/null
> +++ b/lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
> @@ -0,0 +1,85 @@
> +/*-
> + * This file is provided under a dual BSD/GPLv2 license. When using or
> + * redistributing this file, you may do so under either license.
> + *
> + * GPL LICENSE SUMMARY
> + *
> + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of version 2 of the GNU General Public License as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
> + * The full GNU General Public License is included in this distribution
> + * in the file called LICENSE.GPL.
> + *
> + * Contact Information:
> + * Intel Corporation
> + *
> + * BSD LICENSE
> + *
> + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> + * All rights reserved.
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions
> + * are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + * notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + * notice, this list of conditions and the following disclaimer in
> + * the documentation and/or other materials provided with the
> + * distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + * contributors may be used to endorse or promote products derived
> + * from this software without specific prior written permission.
> + *
> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + *
> + */
> +
> +#ifndef _RTE_PCI_DEV_DEFS_H_
> +#define _RTE_PCI_DEV_DEFS_H_
> +
> +#define RTE_PCI_DEV_FEATURE_INTR_MODE "interrupt_mode"
> +
> +#define INTR_NAME_LEN 10
> +
> +#define IGBUIO_NONE_INTR_NAME "none"
> +#define IGBUIO_LEGACY_INTR_NAME "legacy"
> +#define IGBUIO_MSI_INTR_NAME "msi"
> +#define IGBUIO_MSIX_INTR_NAME "msix"
> +
> +
> +#define INTR_MODE(id, mode_name) \
> + .mode = (id), .name = (mode_name)
> +
> +/* interrupt mode */
> +enum igbuio_intr_mode {
> + IGBUIO_NONE_INTR_MODE = 0,
> + IGBUIO_LEGACY_INTR_MODE,
> + IGBUIO_MSI_INTR_MODE,
> + IGBUIO_MSIX_INTR_MODE,
> + IGBUIO_INTR_MODE_MAX
> +};
> +
> +#endif /* _RTE_PCI_DEV_DEFS_H_ */
> diff --git a/lib/librte_eal/common/include/rte_pci_dev_features.h b/lib/librte_eal/common/include/rte_pci_dev_features.h
> new file mode 100755
> index 0000000..a45c056
> --- /dev/null
> +++ b/lib/librte_eal/common/include/rte_pci_dev_features.h
> @@ -0,0 +1,70 @@
> +/*-
> + * This file is provided under a dual BSD/GPLv2 license. When using or
> + * redistributing this file, you may do so under either license.
> + *
> + * GPL LICENSE SUMMARY
> + *
> + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of version 2 of the GNU General Public License as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
> + * The full GNU General Public License is included in this distribution
> + * in the file called LICENSE.GPL.
> + *
> + * Contact Information:
> + * Intel Corporation
> + *
> + * BSD LICENSE
> + *
> + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> + * All rights reserved.
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions
> + * are met:
> + *
> + * * Redistributions of source code must retain the above copyright
> + * notice, this list of conditions and the following disclaimer.
> + * * Redistributions in binary form must reproduce the above copyright
> + * notice, this list of conditions and the following disclaimer in
> + * the documentation and/or other materials provided with the
> + * distribution.
> + * * Neither the name of Intel Corporation nor the names of its
> + * contributors may be used to endorse or promote products derived
> + * from this software without specific prior written permission.
> + *
> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
> + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
> + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
> + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
> + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
> + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
> + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + *
> + */
> +
> +#ifndef RTE_PCI_DEV_INTR_MODE
> +#define RTE_PCI_DEV_INTR_MODE(id, mode_name)
> +#endif
> +
> +RTE_PCI_DEV_INTR_MODE(IGBUIO_NONE_INTR_MODE, IGBUIO_NONE_INTR_NAME)
> +RTE_PCI_DEV_INTR_MODE(IGBUIO_LEGACY_INTR_MODE, IGBUIO_LEGACY_INTR_NAME)
> +RTE_PCI_DEV_INTR_MODE(IGBUIO_MSI_INTR_MODE, IGBUIO_MSI_INTR_NAME)
> +RTE_PCI_DEV_INTR_MODE(IGBUIO_MSIX_INTR_MODE, IGBUIO_MSIX_INTR_NAME)
> +
> +#undef RTE_PCI_DEV_INTR_MODE
> +
> diff --git a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
> index 09c40bf..dd1124b 100644
> --- a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
> +++ b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
> @@ -34,6 +34,8 @@
> #include <xen/xen.h>
> #endif
>
> +#include <rte_pci_dev_feature_defs.h>
> +
> /**
> * MSI-X related macros, copy from linux/pci_regs.h in kernel 2.6.39,
> * but none of them in kernel 2.6.35.
> @@ -49,12 +51,15 @@
>
> #define IGBUIO_NUM_MSI_VECTORS 1
>
> -/* interrupt mode */
> -enum igbuio_intr_mode {
> - IGBUIO_LEGACY_INTR_MODE = 0,
> - IGBUIO_MSI_INTR_MODE,
> - IGBUIO_MSIX_INTR_MODE,
> - IGBUIO_INTR_MODE_MAX
> +struct rte_pci_dev_intr_mode {
> + enum igbuio_intr_mode mode;
> + const char *name;
> +};
> +
> +/* Table of interrupt modes */
> +static struct rte_pci_dev_intr_mode interrupt_modes[] = {
> +#define RTE_PCI_DEV_INTR_MODE(id, mode_name) {INTR_MODE(id, mode_name)},
> +#include <rte_pci_dev_features.h>
> };
>
> /**
> @@ -150,8 +155,32 @@ store_max_vfs(struct device *dev, struct device_attribute *attr,
> }
>
> static DEVICE_ATTR(max_vfs, S_IRUGO | S_IWUSR, show_max_vfs, store_max_vfs);
> +
> +static ssize_t
> +show_interrupt_mode(struct device *dev, struct device_attribute *attr,
> + char *buf)
> +{
> + unsigned i, num_modes;
> + enum igbuio_intr_mode mode;
> + struct uio_info *info;
> + struct pci_dev *pdev = container_of(dev, struct pci_dev, dev);
> +
> + info = pci_get_drvdata(pdev);
> + mode = ((struct rte_uio_pci_dev *)info->priv)->mode;
> + num_modes = sizeof(interrupt_modes)/sizeof(interrupt_modes[0]);
> +
> + for (i = 0; i < num_modes; i++) {
> + if (mode == interrupt_modes[i].mode)
> + return snprintf(buf, INTR_NAME_LEN, "%s\n", interrupt_modes[i].name );
> + }
> + return snprintf(buf, INTR_NAME_LEN, "unknown\n");
> +}
> +
> +static DEVICE_ATTR(interrupt_mode, S_IRUGO, show_interrupt_mode, NULL);
> +
> static struct attribute *dev_attrs[] = {
> &dev_attr_max_vfs.attr,
> + &dev_attr_interrupt_mode.attr,
> NULL,
> };
>
> @@ -520,7 +549,7 @@ igbuio_pci_probe(struct pci_dev *dev, const struct pci_device_id *id)
> #endif
> udev->info.priv = udev;
> udev->pdev = dev;
> - udev->mode = 0; /* set the default value for interrupt mode */
> + udev->mode = IGBUIO_LEGACY_INTR_MODE; /* set the default value for interrupt mode */
> spin_lock_init(&udev->lock);
>
> /* check if it need to try msix first */
> @@ -610,11 +639,10 @@ igbuio_config_intr_mode(char *intr_str)
> printk(KERN_INFO "Use MSIX interrupt by default\n");
> return 0;
> }
> -
> - if (!strcmp(intr_str, "msix")) {
> + if (!strcmp(intr_str, IGBUIO_MSIX_INTR_NAME)) {
> igbuio_intr_mode_preferred = IGBUIO_MSIX_INTR_MODE;
> printk(KERN_INFO "Use MSIX interrupt\n");
> - } else if (!strcmp(intr_str, "legacy")) {
> + } else if (!strcmp(intr_str, IGBUIO_LEGACY_INTR_NAME)) {
> igbuio_intr_mode_preferred = IGBUIO_LEGACY_INTR_MODE;
> printk(KERN_INFO "Use legacy interrupt\n");
> } else {
This assumes that all devices have same interrupt mode.
The whole intr_mode_preferred logic in this driver is borked.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [dpdk-dev] [PATCH 2/4] [PATCH 2/4] eal_pci: Add interrupt mode to rte_pci_device and parsing to eal_pci
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
2014-06-05 14:39 ` [dpdk-dev] [PATCH 1/4] [PATCH 1/4] igb_uio: Add interrupt_mode sysfs entry for igb_uio devices Alan Carew
@ 2014-06-05 14:39 ` Alan Carew
2014-06-05 14:39 ` [dpdk-dev] [PATCH 3/4] [PATCH 3/4] FreeBSD: Adds the equivalent interrupt mode setting and parsing Alan Carew
` (4 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Alan Carew @ 2014-06-05 14:39 UTC (permalink / raw)
To: dev
This patch adds a shared enum between user and kernel space to rte_pci_device.
The value of intr_mode is parsed by eal_pci during pci_uio_map_resource
Signed-off-by: Alan Carew <alan.carew@intel.com>
---
lib/librte_eal/common/Makefile | 1 +
lib/librte_eal/common/include/rte_pci.h | 2 +
lib/librte_eal/linuxapp/eal/eal_pci.c | 78 +++++++++++++++++++++++++++++++
3 files changed, 81 insertions(+), 0 deletions(-)
diff --git a/lib/librte_eal/common/Makefile b/lib/librte_eal/common/Makefile
index 0016fc5..52d46f3 100644
--- a/lib/librte_eal/common/Makefile
+++ b/lib/librte_eal/common/Makefile
@@ -35,6 +35,7 @@ INC := rte_atomic.h rte_branch_prediction.h rte_byteorder.h rte_common.h
INC += rte_cycles.h rte_debug.h rte_eal.h rte_errno.h rte_launch.h rte_lcore.h
INC += rte_log.h rte_memcpy.h rte_memory.h rte_memzone.h rte_pci.h
INC += rte_pci_dev_ids.h rte_per_lcore.h rte_prefetch.h rte_random.h
+INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
INC += rte_rwlock.h rte_spinlock.h rte_tailq.h rte_interrupts.h rte_alarm.h
INC += rte_string_fns.h rte_cpuflags.h rte_version.h rte_tailq_elem.h
INC += rte_eal_memconfig.h rte_malloc_heap.h
diff --git a/lib/librte_eal/common/include/rte_pci.h b/lib/librte_eal/common/include/rte_pci.h
index c793773..5883cd9 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -81,6 +81,7 @@ extern "C" {
#include <stdint.h>
#include <inttypes.h>
#include <rte_interrupts.h>
+#include <rte_pci_dev_feature_defs.h>
TAILQ_HEAD(pci_device_list, rte_pci_device); /**< PCI devices in D-linked Q. */
TAILQ_HEAD(pci_driver_list, rte_pci_driver); /**< PCI drivers in D-linked Q. */
@@ -150,6 +151,7 @@ struct rte_pci_device {
const struct rte_pci_driver *driver; /**< Associated driver */
uint16_t max_vfs; /**< sriov enable if not zero */
int numa_node; /**< NUMA node connection */
+ enum igbuio_intr_mode intr_mode; /**< Interrupt mode */
struct rte_devargs *devargs; /**< Device user arguments */
};
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci.c b/lib/librte_eal/linuxapp/eal/eal_pci.c
index ac2c1fe..7dba446 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci.c
@@ -67,6 +67,7 @@
#include <rte_devargs.h>
#include "rte_pci_dev_ids.h"
+#include "rte_pci_dev_feature_defs.h"
#include "eal_filesystem.h"
#include "eal_private.h"
@@ -89,6 +90,17 @@ struct uio_map {
uint64_t phaddr;
};
+struct rte_pci_dev_intr_mode {
+ enum igbuio_intr_mode mode;
+ const char *name;
+};
+
+/* Table of interrupt modes */
+const struct rte_pci_dev_intr_mode interrupt_modes[] = {
+#define RTE_PCI_DEV_INTR_MODE(id, mode_name) {INTR_MODE(id, mode_name)},
+#include <rte_pci_dev_features.h>
+};
+
/*
* For multi-process we need to reproduce all PCI mappings in secondary
* processes, so save them in a tailq.
@@ -106,6 +118,7 @@ TAILQ_HEAD(uio_res_list, uio_resource);
static struct uio_res_list *uio_res_list = NULL;
static int pci_parse_sysfs_value(const char *filename, uint64_t *val);
+static int pci_parse_sysfs_intr_mode(const char *filename, struct rte_pci_device *dev);
/* unbind kernel driver for this device */
static int
@@ -400,6 +413,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
int i, j;
char dirname[PATH_MAX];
char devname[PATH_MAX]; /* contains the /dev/uioX */
+ char filename[PATH_MAX];
void *mapaddr;
int uio_num;
uint64_t phaddr;
@@ -435,6 +449,18 @@ pci_uio_map_resource(struct rte_pci_device *dev)
}
dev->intr_handle.type = RTE_INTR_HANDLE_UIO;
+ rte_snprintf(filename, sizeof(filename),
+ SYSFS_PCI_DEVICES "/" PCI_PRI_FMT
+ "/"RTE_PCI_DEV_FEATURE_INTR_MODE,
+ loc->domain, loc->bus, loc->devid, loc->function);
+
+ /* Get the kernel configured interrupt mode */
+ if (pci_parse_sysfs_intr_mode(filename, dev) < 0) {
+ RTE_LOG(ERR, EAL, "%s(): cannot determine interrupt_mode\n",
+ __func__);
+ return -1;
+ }
+
/* allocate the mapping details for secondary processes*/
if ((uio_res = rte_zmalloc("UIO_RES", sizeof (*uio_res), 0)) == NULL) {
RTE_LOG(ERR, EAL,
@@ -591,6 +617,58 @@ pci_parse_sysfs_value(const char *filename, uint64_t *val)
return 0;
}
+/*
+ * Parse a sysfs file containing a string
+ */
+static int
+pci_parse_sysfs_string(const char *filename, char *result_str, int max_len)
+{
+ FILE *f;
+ size_t len;
+
+ f = fopen(filename, "r");
+ if (f == NULL) {
+ RTE_LOG(ERR, EAL, "%s(): cannot open sysfs value %s\n",
+ __func__, filename);
+ return -1;
+ }
+ if (fgets(result_str, max_len, f) == NULL) {
+ RTE_LOG(ERR, EAL, "%s(): cannot read sysfs value %s\n",
+ __func__, filename);
+ fclose(f);
+ return -1;
+ }
+ len = strnlen(result_str, max_len) -1;
+ if (result_str[len] == '\n')
+ result_str[len] = '\0';
+
+ fclose(f);
+ return 0;
+}
+
+/*
+ * Determine the kernel configured interrupt mode
+ */
+static int
+pci_parse_sysfs_intr_mode(const char *filename, struct rte_pci_device *dev)
+{
+ char intr_mode[INTR_NAME_LEN];
+ unsigned int i, num_intr_modes = RTE_DIM(interrupt_modes);
+
+ if (pci_parse_sysfs_string(filename, intr_mode, RTE_DIM(intr_mode)) < 0) {
+ RTE_LOG(ERR, EAL, "%s(): cannot parse interrupt_mode\n",
+ __func__);
+ return -1;
+ }
+
+ for (i = 0; i < num_intr_modes; i++)
+ if (!strncmp(intr_mode, interrupt_modes[i].name, INTR_NAME_LEN)) {
+ dev->intr_mode = interrupt_modes[i].mode;
+ return 0;
+ }
+ return -1;
+}
+
/* Compare two PCI device addresses. */
static int
pci_addr_comparison(struct rte_pci_addr *addr, struct rte_pci_addr *addr2)
--
1.7.0.7
^ permalink raw reply [flat|nested] 9+ messages in thread* [dpdk-dev] [PATCH 3/4] [PATCH 3/4] FreeBSD: Adds the equivalent interrupt mode setting and parsing
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
2014-06-05 14:39 ` [dpdk-dev] [PATCH 1/4] [PATCH 1/4] igb_uio: Add interrupt_mode sysfs entry for igb_uio devices Alan Carew
2014-06-05 14:39 ` [dpdk-dev] [PATCH 2/4] [PATCH 2/4] eal_pci: Add interrupt mode to rte_pci_device and parsing to eal_pci Alan Carew
@ 2014-06-05 14:39 ` Alan Carew
2014-06-05 14:39 ` [dpdk-dev] [PATCH 4/4] [PATCH 4/4] virtio: Fixes the VIRTIO_PCI_CONFIG macro to use the correct offset to the Virtio header Alan Carew
` (3 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Alan Carew @ 2014-06-05 14:39 UTC (permalink / raw)
To: dev
This patch adds the equivalent functionality to FreeBSD as with patches 1 and 2
Signed-off-by: Alan Carew <alan.carew@intel.com>
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 45 +++++++++++++++++++++++++++++++
lib/librte_eal/bsdapp/nic_uio/nic_uio.c | 14 +++++++++
2 files changed, 59 insertions(+), 0 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 94ae461..7c270bb 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -50,6 +50,7 @@
#include <sys/ioctl.h>
#include <sys/pciio.h>
#include <dev/pci/pcireg.h>
+#include <sys/sysctl.h>
#include <rte_interrupts.h>
#include <rte_log.h>
@@ -108,6 +109,44 @@ TAILQ_HEAD(uio_res_list, uio_resource);
static struct uio_res_list *uio_res_list = NULL;
+struct rte_pci_dev_intr_mode {
+ enum igbuio_intr_mode mode;
+ const char *name;
+};
+
+/* Table of interrupt modes */
+const struct rte_pci_dev_intr_mode interrupt_modes[] = {
+#define RTE_PCI_DEV_INTR_MODE(id, mode_name) {INTR_MODE (id, mode_name)},
+#include <rte_pci_dev_features.h>
+};
+
+/*
+ * Determine the kernel configured interrupt mode
+ */
+static int
+pci_parse_intr_mode(struct rte_pci_device *dev)
+{
+ char intr_mode[INTR_NAME_LEN];
+ unsigned int i, num_intr_modes = RTE_DIM(interrupt_modes);
+ size_t sysctl_size = sizeof(intr_mode);
+
+ if (sysctlbyname("hw.nic_uio."RTE_PCI_DEV_FEATURE_INTR_MODE, &intr_mode,
+ &sysctl_size, NULL, 0) < 0) {
+ RTE_LOG(ERR, EAL,
+ "%s(): cannot get sysctlbyname: hw.nic_uio.intr_mode\n",
+ __func__);
+ return (-1);
+ }
+ for (i = 0; i < num_intr_modes; i++) {
+ if (!strncmp(intr_mode, interrupt_modes[i].name, INTR_NAME_LEN)) {
+ dev->intr_mode = interrupt_modes[i].mode;
+ return 0;
+ }
+ }
+ return -1;
+
+}
+
/* unbind kernel driver for this device */
static int
pci_unbind_kernel_driver(struct rte_pci_device *dev)
@@ -220,6 +259,12 @@ pci_uio_map_resource(struct rte_pci_device *dev)
return -1;
}
+ if (pci_parse_intr_mode(dev) < 0) {
+ RTE_LOG(ERR, EAL,
+ "%s(): unable to determine interrupt mode\n", __func__);
+ return (-1);
+ }
+
/* save fd if in primary process */
dev->intr_handle.fd = open(devname, O_RDWR);
if (dev->intr_handle.fd < 0) {
diff --git a/lib/librte_eal/bsdapp/nic_uio/nic_uio.c b/lib/librte_eal/bsdapp/nic_uio/nic_uio.c
index c10e9aa..0e17d63 100644
--- a/lib/librte_eal/bsdapp/nic_uio/nic_uio.c
+++ b/lib/librte_eal/bsdapp/nic_uio/nic_uio.c
@@ -51,6 +51,10 @@ __FBSDID("$FreeBSD$");
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/vm_pager.h>
+#include <sys/sysctl.h>
+
+#include <rte_pci_dev_feature_defs.h>
+#include <rte_pci_dev_features.h>
#define MAX_BARS (PCIR_MAX_BAR_0 + 1)
@@ -116,6 +120,16 @@ const struct device devices[] = {
};
#define NUM_DEVICES (sizeof(devices)/sizeof(devices[0]))
+static char nic_uio_intr_mode[] = {IGBUIO_NONE_INTR_NAME};
+
+TUNABLE_STR("hw.nic_uio."RTE_PCI_DEV_FEATURE_INTR_MODE, nic_uio_intr_mode, sizeof(nic_uio_intr_mode));
+
+static SYSCTL_NODE(_hw, OID_AUTO, nic_uio, CTLFLAG_RD, 0, "nic_uio");
+
+SYSCTL_STRING(_hw_nic_uio, OID_AUTO, interrupt_mode, CTLFLAG_RW,
+ &nic_uio_intr_mode, sizeof(nic_uio_intr_mode),
+ "Configured interrupt mode");
+
static devclass_t nic_uio_devclass;
--
1.7.0.7
^ permalink raw reply [flat|nested] 9+ messages in thread* [dpdk-dev] [PATCH 4/4] [PATCH 4/4] virtio: Fixes the VIRTIO_PCI_CONFIG macro to use the correct offset to the Virtio header
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
` (2 preceding siblings ...)
2014-06-05 14:39 ` [dpdk-dev] [PATCH 3/4] [PATCH 3/4] FreeBSD: Adds the equivalent interrupt mode setting and parsing Alan Carew
@ 2014-06-05 14:39 ` Alan Carew
2014-06-05 15:22 ` [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Stephen Hemminger
` (2 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Alan Carew @ 2014-06-05 14:39 UTC (permalink / raw)
To: dev
This final patch address the issue of not being able to determine the correct
offet when MSI-X is disabled.
Signed-off-by: Alan Carew <alan.carew@intel.com>
---
lib/librte_pmd_virtio/virtio_ethdev.c | 1 +
lib/librte_pmd_virtio/virtio_pci.h | 4 +++-
2 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/lib/librte_pmd_virtio/virtio_ethdev.c b/lib/librte_pmd_virtio/virtio_ethdev.c
index c2b4dfb..819e7d7 100644
--- a/lib/librte_pmd_virtio/virtio_ethdev.c
+++ b/lib/librte_pmd_virtio/virtio_ethdev.c
@@ -719,6 +719,7 @@ eth_virtio_dev_init(__rte_unused struct eth_driver *eth_drv,
}
#endif
hw->io_base = (uint32_t)(uintptr_t)pci_dev->mem_resource[0].addr;
+ hw->intr_mode = pci_dev->intr_mode;
/* Reset the device although not necessary at startup */
vtpci_reset(hw);
diff --git a/lib/librte_pmd_virtio/virtio_pci.h b/lib/librte_pmd_virtio/virtio_pci.h
index 96443c7..7b3d0ef 100644
--- a/lib/librte_pmd_virtio/virtio_pci.h
+++ b/lib/librte_pmd_virtio/virtio_pci.h
@@ -44,6 +44,7 @@
#endif
#include <rte_ethdev.h>
+#include <rte_pci_dev_feature_defs.h>
struct virtqueue;
@@ -177,6 +178,7 @@ struct virtio_hw {
uint16_t subsystem_device_id;
uint16_t subsystem_vendor_id;
uint8_t revision_id;
+ enum igbuio_intr_mode intr_mode;
uint8_t mac_addr[ETHER_ADDR_LEN];
int adapter_stopped;
struct rte_eth_stats eth_stats;
@@ -201,7 +203,7 @@ struct virtio_net_config {
* The remaining space is defined by each driver as the per-driver
* configuration space.
*/
-#define VIRTIO_PCI_CONFIG(hw) (((hw)->guest_features & VIRTIO_PCI_FLAG_MSIX) ? 24 : 20)
+#define VIRTIO_PCI_CONFIG(hw) (((hw)->intr_mode == IGBUIO_MSIX_INTR_MODE) ? 24 : 20)
/*
* How many bits to shift physical queue address written to QUEUE_PFN.
--
1.7.0.7
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
` (3 preceding siblings ...)
2014-06-05 14:39 ` [dpdk-dev] [PATCH 4/4] [PATCH 4/4] virtio: Fixes the VIRTIO_PCI_CONFIG macro to use the correct offset to the Virtio header Alan Carew
@ 2014-06-05 15:22 ` Stephen Hemminger
2014-06-05 17:06 ` Stephen Hemminger
2014-06-16 10:19 ` Carew, Alan
6 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2014-06-05 15:22 UTC (permalink / raw)
To: Alan Carew; +Cc: dev
On Thu, 5 Jun 2014 15:39:15 +0100
Alan Carew <alan.carew@intel.com> wrote:
> This series addresses an issue with librte_pmd_virtio where the offset to the
> virtio device specific header may be incorrect depending on whether MSI-X has
> been enabled or not.
>
> If MSI-X is configured the device specific header is placed at byte offset 24
> relative to the IO base address.
> If MSI-X is not configured the device specific header is placed at
> byte offset 20.
>
> The following macro defined in virtio_pci.h is used to test the
> presence of the MSI-X header and determine the correct offset:
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->guest_features & VIRTIO_PCI_FLAG_MSIX) ? 24 : 20)
>
> However, VIRTIO_PCI_FLAG_MSIX is not a guest_feature nor is it part of the
> Virtio Specification and resolves to the VIRTIO_NET_F_MAC feature as both
> are #defined as 0x20.
>
> VIRTIO_PCI_FLAG_MSIX or similar flag should instead be set by the kernel
> driver allocating resources and passed to user space for testing.
> i.e.
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->intr_mode & IGBUIO_MSIX_INTR_MODE) ? 24 : 20)
>
> To enable this testing of interrupt mode, this series allows for the kernel
> driver(igb_uio) to place the configured interrupt mode into a sysfs entry.
> sysfs is then parsed by eal_pci to determine the configured mode, which
> allows all user space devices to correctly determine the interrupt mode,
> including virtio_ethdev.
>
> This series should be applied prior to Anatoly Burakov's
> [VFIO] Add VFIO support to DPDK series
>
> Alan Carew (4):
> igb_uio: Add interrupt_mode sysfs entry for igb_uio devices
> eal_pci: Add interrupt mode to rte_pci_device and parsing to eal_pci
> FreeBSD: Adds the equivalent interrupt mode setting and parsing
> virtio: Fixes the VIRTIO_PCI_CONFIG macro to use the correct offset
> to the Virtio header
>
> lib/librte_eal/bsdapp/eal/eal_pci.c | 44 ++++++++++
> lib/librte_eal/bsdapp/nic_uio/nic_uio.c | 14 +++
> lib/librte_eal/common/Makefile | 1 +
> lib/librte_eal/common/include/rte_pci.h | 2 +
> .../common/include/rte_pci_dev_feature_defs.h | 85 ++++++++++++++++++++
> .../common/include/rte_pci_dev_features.h | 70 ++++++++++++++++
> lib/librte_eal/linuxapp/eal/eal_pci.c | 78 ++++++++++++++++++
> lib/librte_eal/linuxapp/igb_uio/igb_uio.c | 48 +++++++++---
> lib/librte_pmd_virtio/virtio_ethdev.c | 1 +
> lib/librte_pmd_virtio/virtio_pci.h | 4 +-
> 10 files changed, 336 insertions(+), 11 deletions(-)
> create mode 100755 lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
> create mode 100755 lib/librte_eal/common/include/rte_pci_dev_features.h
>
I took the simpler approach. Since DPDK devices doesn't really need the multiple
vectors of MSI-X at all, I just converted igb_uio to use MSI instead.
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
` (4 preceding siblings ...)
2014-06-05 15:22 ` [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Stephen Hemminger
@ 2014-06-05 17:06 ` Stephen Hemminger
2014-06-16 10:19 ` Carew, Alan
6 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2014-06-05 17:06 UTC (permalink / raw)
To: Alan Carew; +Cc: dev
On Thu, 5 Jun 2014 15:39:15 +0100
Alan Carew <alan.carew@intel.com> wrote:
> This series addresses an issue with librte_pmd_virtio where the offset to the
> virtio device specific header may be incorrect depending on whether MSI-X has
> been enabled or not.
>
> If MSI-X is configured the device specific header is placed at byte offset 24
> relative to the IO base address.
> If MSI-X is not configured the device specific header is placed at
> byte offset 20.
>
> The following macro defined in virtio_pci.h is used to test the
> presence of the MSI-X header and determine the correct offset:
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->guest_features & VIRTIO_PCI_FLAG_MSIX) ? 24 : 20)
>
> However, VIRTIO_PCI_FLAG_MSIX is not a guest_feature nor is it part of the
> Virtio Specification and resolves to the VIRTIO_NET_F_MAC feature as both
> are #defined as 0x20.
>
> VIRTIO_PCI_FLAG_MSIX or similar flag should instead be set by the kernel
> driver allocating resources and passed to user space for testing.
> i.e.
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->intr_mode & IGBUIO_MSIX_INTR_MODE) ? 24 : 20)
>
> To enable this testing of interrupt mode, this series allows for the kernel
> driver(igb_uio) to place the configured interrupt mode into a sysfs entry.
> sysfs is then parsed by eal_pci to determine the configured mode, which
> allows all user space devices to correctly determine the interrupt mode,
> including virtio_ethdev.
Why not scan for the IRQ in /proc/interrupts rather than changing the kernel driver.
I ask because my goal is to get rid of igb_uio and it's quirks and use
uio_pci_generic instead.
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested
2014-06-05 14:39 [dpdk-dev] [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio feature bit-flag tested Alan Carew
` (5 preceding siblings ...)
2014-06-05 17:06 ` Stephen Hemminger
@ 2014-06-16 10:19 ` Carew, Alan
6 siblings, 0 replies; 9+ messages in thread
From: Carew, Alan @ 2014-06-16 10:19 UTC (permalink / raw)
To: dev
> -----Original Message-----
> From: Carew, Alan
> Sent: Thursday, June 05, 2014 3:39 PM
> To: dev@dpdk.org
> Cc: Carew, Alan
> Subject: [PATCH 0/4] librte_pmd_virtio :Fix: virtio_pci.h non-existent virtio
> feature bit-flag tested
>
> This series addresses an issue with librte_pmd_virtio where the offset to the
> virtio device specific header may be incorrect depending on whether MSI-X has
> been enabled or not.
>
> If MSI-X is configured the device specific header is placed at byte offset 24
> relative to the IO base address.
> If MSI-X is not configured the device specific header is placed at
> byte offset 20.
>
> The following macro defined in virtio_pci.h is used to test the
> presence of the MSI-X header and determine the correct offset:
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->guest_features &
> VIRTIO_PCI_FLAG_MSIX) ? 24 : 20)
>
> However, VIRTIO_PCI_FLAG_MSIX is not a guest_feature nor is it part of the
> Virtio Specification and resolves to the VIRTIO_NET_F_MAC feature as both
> are #defined as 0x20.
>
> VIRTIO_PCI_FLAG_MSIX or similar flag should instead be set by the kernel
> driver allocating resources and passed to user space for testing.
> i.e.
> #define VIRTIO_PCI_CONFIG(hw) (((hw)->intr_mode &
> IGBUIO_MSIX_INTR_MODE) ? 24 : 20)
>
> To enable this testing of interrupt mode, this series allows for the kernel
> driver(igb_uio) to place the configured interrupt mode into a sysfs entry.
> sysfs is then parsed by eal_pci to determine the configured mode, which
> allows all user space devices to correctly determine the interrupt mode,
> including virtio_ethdev.
>
> This series should be applied prior to Anatoly Burakov's
> [VFIO] Add VFIO support to DPDK series
>
> Alan Carew (4):
> igb_uio: Add interrupt_mode sysfs entry for igb_uio devices
> eal_pci: Add interrupt mode to rte_pci_device and parsing to eal_pci
> FreeBSD: Adds the equivalent interrupt mode setting and parsing
> virtio: Fixes the VIRTIO_PCI_CONFIG macro to use the correct offset
> to the Virtio header
>
> lib/librte_eal/bsdapp/eal/eal_pci.c | 44 ++++++++++
> lib/librte_eal/bsdapp/nic_uio/nic_uio.c | 14 +++
> lib/librte_eal/common/Makefile | 1 +
> lib/librte_eal/common/include/rte_pci.h | 2 +
> .../common/include/rte_pci_dev_feature_defs.h | 85
> ++++++++++++++++++++
> .../common/include/rte_pci_dev_features.h | 70 ++++++++++++++++
> lib/librte_eal/linuxapp/eal/eal_pci.c | 78 ++++++++++++++++++
> lib/librte_eal/linuxapp/igb_uio/igb_uio.c | 48 +++++++++---
> lib/librte_pmd_virtio/virtio_ethdev.c | 1 +
> lib/librte_pmd_virtio/virtio_pci.h | 4 +-
> 10 files changed, 336 insertions(+), 11 deletions(-)
> create mode 100755
> lib/librte_eal/common/include/rte_pci_dev_feature_defs.h
> create mode 100755 lib/librte_eal/common/include/rte_pci_dev_features.h
Self NAK, alternative approach and patch set to follow.
^ permalink raw reply [flat|nested] 9+ messages in thread