From: Jerome Glisse <j.glisse@gmail.com>
To: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Andrew Lewycky <Andrew.Lewycky@amd.com>,
Ben Goz <ben.goz@amd.com>,
linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org,
Evgeny Pinchuk <evgeny.pinchuk@amd.com>,
Alexey Skidanov <Alexey.Skidanov@amd.com>,
linux-api@vger.kernel.org,
Alex Deucher <alexander.deucher@amd.com>
Subject: Re: [PATCH 32/83] hsa/radeon: implementing IOCTL for clock counters
Date: Fri, 11 Jul 2014 16:34:56 -0400 [thread overview]
Message-ID: <20140711203455.GR1870@gmail.com> (raw)
In-Reply-To: <1405029279-6894-4-git-send-email-oded.gabbay@amd.com>
On Fri, Jul 11, 2014 at 12:53:48AM +0300, Oded Gabbay wrote:
> From: Evgeny Pinchuk <evgeny.pinchuk@amd.com>
>
> Implemented new IOCTL to query the CPU and GPU clock counters.
>
> Signed-off-by: Evgeny Pinchuk <evgeny.pinchuk@amd.com>
> Signed-off-by: Oded Gabbay <oded.gabbay@amd.com>
> ---
> drivers/gpu/hsa/radeon/kfd_chardev.c | 37 ++++++++++++++++++++++++++++++++++++
> include/uapi/linux/kfd_ioctl.h | 9 +++++++++
> 2 files changed, 46 insertions(+)
>
> diff --git a/drivers/gpu/hsa/radeon/kfd_chardev.c b/drivers/gpu/hsa/radeon/kfd_chardev.c
> index ddaf357..d6fa980 100644
> --- a/drivers/gpu/hsa/radeon/kfd_chardev.c
> +++ b/drivers/gpu/hsa/radeon/kfd_chardev.c
> @@ -28,6 +28,7 @@
> #include <linux/slab.h>
> #include <linux/uaccess.h>
> #include <uapi/linux/kfd_ioctl.h>
> +#include <linux/time.h>
> #include "kfd_priv.h"
> #include "kfd_scheduler.h"
>
> @@ -284,6 +285,38 @@ out:
> return err;
> }
>
> +static long
> +kfd_ioctl_get_clock_counters(struct file *filep, struct kfd_process *p, void __user *arg)
> +{
> + struct kfd_ioctl_get_clock_counters_args args;
> + struct kfd_dev *dev;
> + struct timespec time;
> +
> + if (copy_from_user(&args, arg, sizeof(args)))
> + return -EFAULT;
> +
> + dev = radeon_kfd_device_by_id(args.gpu_id);
> + if (dev == NULL)
> + return -EINVAL;
> +
> + /* Reading GPU clock counter from KGD */
> + args.gpu_clock_counter = kfd2kgd->get_gpu_clock_counter(dev->kgd);
> +
> + /* No access to rdtsc. Using raw monotonic time */
> + getrawmonotonic(&time);
> + args.cpu_clock_counter = time.tv_nsec;
Is the GPU clock counter monotonic too ? Even after GPU reset (hard reset
included) what could go wrong if it rolls back ?
> +
> + get_monotonic_boottime(&time);
> + args.system_clock_counter = time.tv_nsec;
> +
> + /* Since the counter is in nano-seconds we use 1GHz frequency */
> + args.system_clock_freq = 1000000000;
> +
> + if (copy_to_user(arg, &args, sizeof(args)))
> + return -EFAULT;
> +
> + return 0;
> +}
>
> static long
> kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
> @@ -312,6 +345,10 @@ kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
> err = kfd_ioctl_set_memory_policy(filep, process, (void __user *)arg);
> break;
>
> + case KFD_IOC_GET_CLOCK_COUNTERS:
> + err = kfd_ioctl_get_clock_counters(filep, process, (void __user *)arg);
> + break;
> +
> default:
> dev_err(kfd_device,
> "unknown ioctl cmd 0x%x, arg 0x%lx)\n",
> diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h
> index 928e628..5b9517e 100644
> --- a/include/uapi/linux/kfd_ioctl.h
> +++ b/include/uapi/linux/kfd_ioctl.h
> @@ -70,12 +70,21 @@ struct kfd_ioctl_set_memory_policy_args {
> uint64_t alternate_aperture_size; /* to KFD */
> };
>
> +struct kfd_ioctl_get_clock_counters_args {
> + uint32_t gpu_id; /* to KFD */
> + uint64_t gpu_clock_counter; /* from KFD */
> + uint64_t cpu_clock_counter; /* from KFD */
> + uint64_t system_clock_counter; /* from KFD */
> + uint64_t system_clock_freq; /* from KFD */
> +};
> +
> #define KFD_IOC_MAGIC 'K'
>
> #define KFD_IOC_GET_VERSION _IOR(KFD_IOC_MAGIC, 1, struct kfd_ioctl_get_version_args)
> #define KFD_IOC_CREATE_QUEUE _IOWR(KFD_IOC_MAGIC, 2, struct kfd_ioctl_create_queue_args)
> #define KFD_IOC_DESTROY_QUEUE _IOWR(KFD_IOC_MAGIC, 3, struct kfd_ioctl_destroy_queue_args)
> #define KFD_IOC_SET_MEMORY_POLICY _IOW(KFD_IOC_MAGIC, 4, struct kfd_ioctl_set_memory_policy_args)
> +#define KFD_IOC_GET_CLOCK_COUNTERS _IOWR(KFD_IOC_MAGIC, 5, struct kfd_ioctl_get_clock_counters_args)
>
> #pragma pack(pop)
>
> --
> 1.9.1
>
next prev parent reply other threads:[~2014-07-11 20:34 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1405029279-6894-1-git-send-email-oded.gabbay@amd.com>
[not found] ` <1405029279-6894-1-git-send-email-oded.gabbay-5C7GfCeVMHo@public.gmane.org>
2014-07-10 21:53 ` [PATCH 32/83] hsa/radeon: implementing IOCTL for clock counters Oded Gabbay
2014-07-11 20:34 ` Jerome Glisse [this message]
2014-07-10 21:53 ` [PATCH 42/83] hsa/radeon: 32-bit processes support Oded Gabbay
2014-07-10 21:54 ` [PATCH 44/83] hsa/radeon: HSA64/HSA32 modes support Oded Gabbay
[not found] ` <1405029279-6894-16-git-send-email-oded.gabbay-5C7GfCeVMHo@public.gmane.org>
2014-07-11 20:41 ` Jerome Glisse
2014-07-10 21:54 ` [PATCH 54/83] hsa/radeon: Switch to new queue scheduler Oded Gabbay
2014-07-10 21:54 ` [PATCH 55/83] hsa/radeon: Add IOCTL for update queue Oded Gabbay
2014-07-10 21:54 ` [PATCH 59/83] hsa/radeon: Exclusive access for perf. counters Oded Gabbay
2014-07-10 21:54 ` [PATCH 60/83] hsa/radeon: Rearrange structures in kfd_ioctl.h Oded Gabbay
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=20140711203455.GR1870@gmail.com \
--to=j.glisse@gmail.com \
--cc=Alexey.Skidanov@amd.com \
--cc=Andrew.Lewycky@amd.com \
--cc=alexander.deucher@amd.com \
--cc=ben.goz@amd.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=evgeny.pinchuk@amd.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=oded.gabbay@gmail.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).