Re: [PATCH v9 3/5] hisi_acc_vfio_pci: register debugfs for hisilicon migration driver

From: Alex Williamson
Date: Mon Mar 20 2023 - 18:10:00 EST

Next message: Aleksa Sarai: "Re: [PATCH] do_open(): Fix O_DIRECTORY | O_CREAT behavior"
Previous message: Willy Tarreau: "Re: [PATCH v2 0/8] tools/nolibc: add support for stack protector"
In reply to: Longfang Liu: "[PATCH v9 3/5] hisi_acc_vfio_pci: register debugfs for hisilicon migration driver"
Next in thread: Longfang Liu: "[PATCH v9 4/5] Documentation: add debugfs description for vfio"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Mon, 20 Mar 2023 17:23:36 +0800
Longfang Liu <liulongfang@xxxxxxxxxx> wrote:

> On the debugfs framework of VFIO, if the CONFIG_DEBUG_FS macro is
> enabled, the debug function is registered for the live migration driver
> of the HiSilicon accelerator device.
>
> After registering the HiSilicon accelerator device on the debugfs
> framework of live migration of vfio, a directory file "hisi_acc"
> of debugfs is created, and then three debug function files are
> created in this directory:
>
> data file: used to get the migration data of the live migration device
> attr file: used to get device attributes of the live migration device
> debug file: used to test for acquiring and writing device state data
> for VF device.
>
> Signed-off-by: Longfang Liu <liulongfang@xxxxxxxxxx>
> ---
> .../vfio/pci/hisilicon/hisi_acc_vfio_pci.c | 290 ++++++++++++++++++
> .../vfio/pci/hisilicon/hisi_acc_vfio_pci.h | 11 +
> 2 files changed, 301 insertions(+)
>
> diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> index a1589947e721..22ebe0e8f90f 100644
> --- a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> +++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> @@ -15,6 +15,7 @@
> #include <linux/anon_inodes.h>
>
> #include "hisi_acc_vfio_pci.h"
> +#include "../../vfio.h"
>
> /* Return 0 on VM acc device ready, -ETIMEDOUT hardware timeout */
> static int qm_wait_dev_not_ready(struct hisi_qm *qm)
> @@ -606,6 +607,18 @@ hisi_acc_check_int_state(struct hisi_acc_vf_core_device *hisi_acc_vdev)
> }
> }
>
> +static void hisi_acc_vf_migf_save(struct hisi_acc_vf_migration_file *src_migf,
> + struct hisi_acc_vf_migration_file *dst_migf)
> +{
> + if (!dst_migf)
> + return;
> +
> + dst_migf->disabled = false;
> + dst_migf->total_length = src_migf->total_length;
> + memcpy(&dst_migf->vf_data, &src_migf->vf_data,
> + sizeof(struct acc_vf_data));
> +}
> +
> static void hisi_acc_vf_disable_fd(struct hisi_acc_vf_migration_file *migf)
> {
> mutex_lock(&migf->lock);
> @@ -618,12 +631,16 @@ static void hisi_acc_vf_disable_fd(struct hisi_acc_vf_migration_file *migf)
> static void hisi_acc_vf_disable_fds(struct hisi_acc_vf_core_device *hisi_acc_vdev)
> {
> if (hisi_acc_vdev->resuming_migf) {
> + hisi_acc_vf_migf_save(hisi_acc_vdev->resuming_migf,
> + hisi_acc_vdev->debug_migf);
> hisi_acc_vf_disable_fd(hisi_acc_vdev->resuming_migf);
> fput(hisi_acc_vdev->resuming_migf->filp);
> hisi_acc_vdev->resuming_migf = NULL;
> }
>
> if (hisi_acc_vdev->saving_migf) {
> + hisi_acc_vf_migf_save(hisi_acc_vdev->saving_migf,
> + hisi_acc_vdev->debug_migf);
> hisi_acc_vf_disable_fd(hisi_acc_vdev->saving_migf);
> fput(hisi_acc_vdev->saving_migf->filp);
> hisi_acc_vdev->saving_migf = NULL;
> @@ -1303,6 +1320,273 @@ static long hisi_acc_vfio_pci_ioctl(struct vfio_device *core_vdev, unsigned int
> return vfio_pci_core_ioctl(core_vdev, cmd, arg);
> }
>
> +static int hisi_acc_vf_debug_io(struct vfio_device *vdev)
> +{
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_qm *vf_qm = &hisi_acc_vdev->vf_qm;
> + struct device *dev = vdev->dev;
> + u64 data;
> + int ret;
> +
> + ret = qm_wait_dev_not_ready(vf_qm);
> + if (ret)
> + dev_err(dev, "VF device not ready!\n");
> +
> + data = readl(vf_qm->io_base + QM_MB_CMD_SEND_BASE);
> + dev_info(dev, "debug mailbox val: 0x%llx\n", data);
> +
> + return ret;
> +}
> +
> +static int hisi_acc_vf_debug_resume(struct vfio_device *vdev)
> +{
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_acc_vf_migration_file *migf = hisi_acc_vdev->debug_migf;
> + struct device *dev = vdev->dev;
> + int ret;
> +
> + ret = vf_qm_state_save(hisi_acc_vdev, migf);
> + if (ret) {
> + dev_err(dev, "failed to save device data!\n");
> + return -EINVAL;
> + }
> +
> + ret = vf_qm_check_match(hisi_acc_vdev, migf);
> + if (ret) {
> + dev_err(dev, "failed to match the VF!\n");
> + return -EINVAL;
> + }
> +
> + ret = vf_qm_load_data(hisi_acc_vdev, migf);
> + if (ret) {
> + dev_err(dev, "failed to recover the VF!\n");
> + return -EINVAL;
> + }
> +
> + vf_qm_fun_reset(&hisi_acc_vdev->vf_qm);
> + dev_info(dev, "successful to resume device data!\n");
> +
> + return 0;
> +}
> +
> +static int hisi_acc_vf_debug_save(struct vfio_device *vdev)
> +{
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_acc_vf_migration_file *migf = hisi_acc_vdev->debug_migf;
> + struct device *dev = vdev->dev;
> + int ret;
> +
> + ret = vf_qm_state_save(hisi_acc_vdev, migf);
> + if (ret) {
> + dev_err(dev, "failed to save device data!\n");
> + return -EINVAL;
> + }
> + dev_info(dev, "successful to save device data!\n");
> +
> + return 0;
> +}

It seems backwards that we've defined the 'debug' file to be writable
only in order to select which operation to perform, where the user then
needs to look in dmesg for information related to an errno. Wouldn't
it make more sense to create separate save, restore, and io_test files
that perform the operation when read and return a buffer indicating the
error condition or success? What's the benefit of multiplexing these
through writes to a single debug file?

> +
> +static int
> +hisi_acc_vf_debug_operate(struct vfio_device *vdev, unsigned int cmd)
> +{
> + int ret;
> +
> + switch (cmd) {
> + case STATE_SAVE:
> + ret = hisi_acc_vf_debug_save(vdev);
> + if (ret)
> + return ret;
> + break;
> + case STATE_RESUME:
> + ret = hisi_acc_vf_debug_resume(vdev);
> + if (ret)
> + return ret;
> + break;
> + case RW_IO_TEST:
> + ret = hisi_acc_vf_debug_io(vdev);
> + if (ret)
> + return ret;
> + break;
> + default:
> + return -EINVAL;
> + }
> +
> + return 0;
> +}

Initialize ret = -EINVAL, 'return ret' outside the case, and drop the
per-case returns.

> +
> +static int hisi_acc_vf_debug_check(struct vfio_device *vdev)
> +{
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_acc_vf_migration_file *migf = hisi_acc_vdev->debug_migf;
> + enum vfio_device_mig_state state;
> +
> + if (!vdev->mig_ops || !migf) {
> + dev_err(vdev->dev, "device not support debugfs!\n");
> + return -EINVAL;
> + }

Why did we create the file then?

> +
> + /* If device not opened, the debugfs operation will trigger calltrace */
> + (void)vdev->mig_ops->migration_get_state(vdev, &state);
> + if (state == VFIO_DEVICE_STATE_ERROR ||
> + state == VFIO_DEVICE_STATE_STOP) {
> + dev_err(vdev->dev, "device not opened!\n");
> + return -EINVAL;
> + }

And we can't use something like open_count? What protects this from
racing against the device being released?

> +
> + return 0;
> +}
> +
> +static ssize_t hisi_acc_vf_debug_write(struct file *filp, const char __user *buffer,
> + size_t count, loff_t *pos)
> +{
> + struct vfio_device *vdev = filp->private_data;
> + char tbuf[VFIO_DEV_DBG_LEN];
> + unsigned long cmd;
> + int len, ret;
> +
> + if (*pos)
> + return 0;
> +
> + if (count >= VFIO_DEV_DBG_LEN)
> + return -ENOSPC;
> +
> + ret = hisi_acc_vf_debug_check(vdev);
> + if (ret)
> + return -EINVAL;
> +
> + len = simple_write_to_buffer(tbuf, VFIO_DEV_DBG_LEN - 1,
> + pos, buffer, count);
> + if (len < 0 || len > VFIO_DEV_DBG_LEN - 1)
> + return -EINVAL;
> + tbuf[len] = '\0';
> + if (kstrtoul(tbuf, 0, &cmd))
> + return -EFAULT;
> +
> + ret = hisi_acc_vf_debug_operate(vdev, cmd);
> + if (ret) {
> + dev_err(vdev->dev, "device debug test failed!\n");
> + return -EINVAL;
> + }
> +
> + return count;
> +}
> +
> +static const struct file_operations hisi_acc_vf_debug_fops = {
> + .owner = THIS_MODULE,
> + .open = simple_open,
> + .write = hisi_acc_vf_debug_write,
> +};
> +
> +static void seq_print_hex_data(struct seq_file *seq, const void *buf, size_t len)
> +{
> +#define HEX_LINE_SIZE 131
> +#define HEX_ROW_SIZE 16
> + unsigned char linebuf[HEX_LINE_SIZE];
> + int i, linelen, remaining = len;
> + const u8 *ptr = buf;
> +
> + for (i = 0; i < len; i += HEX_ROW_SIZE) {
> + linelen = min(remaining, HEX_ROW_SIZE);
> + remaining -= HEX_ROW_SIZE;
> +
> + hex_dump_to_buffer(ptr + i, linelen, HEX_ROW_SIZE, 1,
> + linebuf, sizeof(linebuf), false);
> +
> + seq_printf(seq, "%s%.8x: %s\n", "Mig Data:", i, linebuf);
> + }
> +}

Can't this use seq_hex_dump()? If there's a _good_ reason why not, it
should be noted in comments and the function should be renamed not to
pollute the seq_ namespace. Thanks,

Alex

> +
> +static int hisi_acc_vf_data_read(struct seq_file *seq, void *data)
> +{
> + struct device *vf_dev = seq->private;
> + struct vfio_pci_core_device *core_device = dev_get_drvdata(vf_dev);
> + struct vfio_device *vdev = &core_device->vdev;
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_acc_vf_migration_file *debug_migf = hisi_acc_vdev->debug_migf;
> + size_t vf_data_sz = offsetofend(struct acc_vf_data, padding);
> +
> + if (debug_migf && debug_migf->total_length)
> + seq_print_hex_data(seq, (unsigned char *)&debug_migf->vf_data,
> + vf_data_sz);
> + else
> + seq_printf(seq, "%s\n", "device not migrated!");
> +
> + return 0;
> +}
> +
> +static int hisi_acc_vf_attr_read(struct seq_file *seq, void *data)
> +{
> + struct device *vf_dev = seq->private;
> + struct vfio_pci_core_device *core_device = dev_get_drvdata(vf_dev);
> + struct vfio_device *vdev = &core_device->vdev;
> + struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(vdev);
> + struct hisi_acc_vf_migration_file *debug_migf = hisi_acc_vdev->debug_migf;
> +
> + if (debug_migf && debug_migf->total_length) {
> + seq_printf(seq,
> + "acc device:\n"
> + "device state: %d\n"
> + "device ready: %u\n"
> + "data valid: %d\n"
> + "data size: %lu\n",
> + hisi_acc_vdev->mig_state,
> + hisi_acc_vdev->vf_qm_state,
> + debug_migf->disabled,
> + debug_migf->total_length);
> + } else {
> + seq_printf(seq, "%s\n", "device not migrated!");
> + }
> +
> + return 0;
> +}
> +
> +static int hisi_acc_vfio_debug_init(struct hisi_acc_vf_core_device *hisi_acc_vdev)
> +{
> + struct vfio_device *vdev = &hisi_acc_vdev->core_device.vdev;
> + struct dentry *vfio_dev_migration = NULL;
> + struct dentry *vfio_hisi_acc = NULL;
> + struct device *dev = vdev->dev;
> + void *migf = NULL;
> +
> + if (!debugfs_initialized())
> + return 0;
> +
> + vfio_device_debugfs_init(vdev);
> +
> + migf = kzalloc(sizeof(struct hisi_acc_vf_migration_file), GFP_KERNEL);
> + if (!migf)
> + return -ENOMEM;
> +
> + hisi_acc_vdev->debug_migf = migf;
> +
> + vfio_dev_migration = debugfs_lookup("migration", vdev->debug_root);
> + if (!vfio_dev_migration) {
> + dev_err(dev, "failed to lookup migration debugfs file!\n");
> + return -ENODEV;
> + }
> +
> + vfio_hisi_acc = debugfs_create_dir("hisi_acc", vfio_dev_migration);
> + debugfs_create_devm_seqfile(dev, "data", vfio_hisi_acc,
> + hisi_acc_vf_data_read);
> + debugfs_create_devm_seqfile(dev, "attr", vfio_hisi_acc,
> + hisi_acc_vf_attr_read);
> + debugfs_create_file("debug", 0200, vfio_hisi_acc,
> + vdev, &hisi_acc_vf_debug_fops);
> +
> + return 0;
> +}
> +
> +static void hisi_acc_vf_debugfs_exit(struct hisi_acc_vf_core_device *hisi_acc_vdev)
> +{
> + struct vfio_device *vdev = &hisi_acc_vdev->core_device.vdev;
> +
> + if (!debugfs_initialized())
> + return;
> +
> + vfio_device_debugfs_exit(vdev);
> + kfree(hisi_acc_vdev->debug_migf);
> +}
> +
> static int hisi_acc_vfio_pci_open_device(struct vfio_device *core_vdev)
> {
> struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_get_vf_dev(core_vdev);
> @@ -1416,6 +1700,9 @@ static int hisi_acc_vfio_pci_probe(struct pci_dev *pdev, const struct pci_device
> if (IS_ERR(hisi_acc_vdev))
> return PTR_ERR(hisi_acc_vdev);
>
> + if (ops == &hisi_acc_vfio_pci_migrn_ops)
> + hisi_acc_vfio_debug_init(hisi_acc_vdev);
> +
> dev_set_drvdata(&pdev->dev, &hisi_acc_vdev->core_device);
> ret = vfio_pci_core_register_device(&hisi_acc_vdev->core_device);
> if (ret)
> @@ -1423,6 +1710,8 @@ static int hisi_acc_vfio_pci_probe(struct pci_dev *pdev, const struct pci_device
> return 0;
>
> out_put_vdev:
> + if (ops == &hisi_acc_vfio_pci_migrn_ops)
> + hisi_acc_vf_debugfs_exit(hisi_acc_vdev);
> vfio_put_device(&hisi_acc_vdev->core_device.vdev);
> return ret;
> }
> @@ -1431,6 +1720,7 @@ static void hisi_acc_vfio_pci_remove(struct pci_dev *pdev)
> {
> struct hisi_acc_vf_core_device *hisi_acc_vdev = hisi_acc_drvdata(pdev);
>
> + hisi_acc_vf_debugfs_exit(hisi_acc_vdev);
> vfio_pci_core_unregister_device(&hisi_acc_vdev->core_device);
> vfio_put_device(&hisi_acc_vdev->core_device.vdev);
> }
> diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.h b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.h
> index dcabfeec6ca1..ef50b12f018d 100644
> --- a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.h
> +++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.h
> @@ -49,6 +49,14 @@
> #define QM_EQC_DW0 0X8000
> #define QM_AEQC_DW0 0X8020
>
> +#define VFIO_DEV_DBG_LEN 256
> +
> +enum mig_debug_cmd {
> + STATE_SAVE,
> + STATE_RESUME,
> + RW_IO_TEST,
> +};
> +
> struct acc_vf_data {
> #define QM_MATCH_SIZE offsetofend(struct acc_vf_data, qm_rsv_state)
> /* QM match information */
> @@ -113,5 +121,8 @@ struct hisi_acc_vf_core_device {
> spinlock_t reset_lock;
> struct hisi_acc_vf_migration_file *resuming_migf;
> struct hisi_acc_vf_migration_file *saving_migf;
> +
> + /* For debugfs */
> + struct hisi_acc_vf_migration_file *debug_migf;
> };
> #endif /* HISI_ACC_VFIO_PCI_H */

Next message: Aleksa Sarai: "Re: [PATCH] do_open(): Fix O_DIRECTORY | O_CREAT behavior"
Previous message: Willy Tarreau: "Re: [PATCH v2 0/8] tools/nolibc: add support for stack protector"
In reply to: Longfang Liu: "[PATCH v9 3/5] hisi_acc_vfio_pci: register debugfs for hisilicon migration driver"
Next in thread: Longfang Liu: "[PATCH v9 4/5] Documentation: add debugfs description for vfio"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]