dri-devel September 2014

dri-devel@lists.freedesktop.org

173 participants
1045 discussions

[Bug 22312] radeon 0000:01:00.0: Invalid ROM contents
by bugzilla-daemon＠bugzilla.kernel.org 27 Sep '14

27 Sep '14

https://bugzilla.kernel.org/show_bug.cgi?id=22312 Jarkko K <jarkko_korpi(a)hotmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jarkko_korpi(a)hotmail.com --- Comment #11 from Jarkko K <jarkko_korpi(a)hotmail.com> --- I also have the error and and suddenly these lines [ 4985.275572] radeon 0000:01:00.0: GPU fault detected: 146 0x0f0a8804 [ 4985.275573] radeon 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00009678 [ 4985.275574] radeon 0000:01:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x0A088004 [ 4985.275576] VM fault (0x04, vmid 5) at page 38520, read from 'TC4' (0x54433400) (136) r9 290 3.16.6rc kernel -- You are receiving this mail because: You are watching the assignee of the bug.

1 0

[PATCH RFC v3] vga_switcheroo: restore mux and power state upon resume from hibernation
by Stefan Demharter 26 Sep '14

26 Sep '14

Restore state of mux switch and power-off state of graphics cards upon resume from hibernation as the following problems can occur: If the mux is set to the discrete card prior to hibernation it is reset to the integrated card by the system upon boot. Thus it stays at the wrong state after resume which results in a black screen. Solve this by saving the current state of the mux and restoring it upon resume from hibernation. If a graphics card is powered off prior to hibernation it is powered up by the system upon boot. Thus it stays still active after resume and there is an inconsistence between the power state the kernel thinks the graphic card is in and the actual power state. In case of the discrete card this also results in a noticeable increase in power consumption. Solve this by powering off such cards upon resume. Signed-off-by: Stefan Demharter <stefan.demharter(a)gmx.net> --- drivers/gpu/vga/vga_switcheroo.c | 123 +++++++++++++++++++++++++++++++++++---- 1 file changed, 112 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/vga/vga_switcheroo.c b/drivers/gpu/vga/vga_switcheroo.c index 37ac7b5..67a82f1 100644 --- a/drivers/gpu/vga/vga_switcheroo.c +++ b/drivers/gpu/vga/vga_switcheroo.c @@ -28,6 +28,7 @@ #include <linux/console.h> #include <linux/vga_switcheroo.h> #include <linux/pm_runtime.h> +#include <linux/suspend.h> #include <linux/vgaarb.h> @@ -72,6 +73,89 @@ static struct vgasr_priv vgasr_priv = { .clients = LIST_HEAD_INIT(vgasr_priv.clients), }; +static int vga_switchoff(struct vga_switcheroo_client *client); + +/* Assume that the initial state of the mux is set to IGD at boot time */ +#define MUX_INITIAL_STATE VGA_SWITCHEROO_IGD + +/* + * The next variable stores the state of the mux + * so that it can be restored upon resume from hibernation. + * + * This should also handle muxless systems quite well + * since there mux_state is never altered and the function + * restore_mux will end up doing nothing + */ +static int mux_state = MUX_INITIAL_STATE; + +#define DIS_IGD_STR(id) ((id) == VGA_SWITCHEROO_DIS ? "DIS" : "IGD") + +static int vga_switchon(struct vga_switcheroo_client *client); +static int vga_switchoff(struct vga_switcheroo_client *client); + +/* Restore saved mux state */ +static void vga_restore_mux(void) +{ + if (vgasr_priv.handler->switchto && + mux_state != MUX_INITIAL_STATE) { + int ret; + + pr_info("vga_switcheroo: restoring mux state to %s\n", + DIS_IGD_STR(mux_state)); + ret = vgasr_priv.handler->switchto(mux_state); + + if (ret) + pr_warn("vga_switcheroo: failed (%d)\n", ret); + } +} + +/* + * Power off devices from which the system thinks they are powered-off. + * If a device is powered-off before hibernation the system may + * power it on upon boot. This results in a mismatch between what + * the driver thinks of the power state and the actual power state. + * This function resolves this mismatch by powering-off corresponding devices. + */ +static void vga_restore_power_off_state(void) +{ + struct vga_switcheroo_client *client; + + list_for_each_entry(client, &vgasr_priv.clients, list) { + if (client_is_audio(client)) + continue; + + if (client->pwr_state == VGA_SWITCHEROO_OFF) { + pr_info("vga_switcheroo: Restoring power-off state for %s\n", + DIS_IGD_STR(client->id)); + vgasr_priv.handler->power_state(client->id, VGA_SWITCHEROO_OFF); + } + } +} + +/* + * Callback to handle resume from hibernation. + * Restore mux state and power states for switcheroo controlled GPUs. + */ +static int vga_switcheroo_resume(struct notifier_block *nb, + unsigned long action, void *unused) +{ + if (action == PM_POST_HIBERNATION) { + mutex_lock(&vgasr_mutex); + + vga_restore_mux(); + vga_restore_power_off_state(); + + mutex_unlock(&vgasr_mutex); + } + + return 0; +} + +static struct notifier_block vga_pm_nb = { + .notifier_call = vga_switcheroo_resume, + .priority = 0, +}; + static bool vga_switcheroo_ready(void) { /* we're ready if we get two clients + handler */ @@ -83,6 +167,7 @@ static void vga_switcheroo_enable(void) { int ret; struct vga_switcheroo_client *client; + register_pm_notifier(&vga_pm_nb); /* call the handler to init */ if (vgasr_priv.handler->init) @@ -119,14 +204,20 @@ int vga_switcheroo_register_handler(struct vga_switcheroo_handler *handler) } EXPORT_SYMBOL(vga_switcheroo_register_handler); +void vga_switcheroo_disable(void) +{ + unregister_pm_notifier(&vga_pm_nb); + pr_info("vga_switcheroo: disabled\n"); + vga_switcheroo_debugfs_fini(&vgasr_priv); + vgasr_priv.active = false; +} + void vga_switcheroo_unregister_handler(void) { mutex_lock(&vgasr_mutex); vgasr_priv.handler = NULL; if (vgasr_priv.active) { - pr_info("vga_switcheroo: disabled\n"); - vga_switcheroo_debugfs_fini(&vgasr_priv); - vgasr_priv.active = false; + vga_switcheroo_disable(); } mutex_unlock(&vgasr_mutex); } @@ -235,9 +326,7 @@ void vga_switcheroo_unregister_client(struct pci_dev *pdev) kfree(client); } if (vgasr_priv.active && vgasr_priv.registered_clients < 2) { - printk(KERN_INFO "vga_switcheroo: disabled\n"); - vga_switcheroo_debugfs_fini(&vgasr_priv); - vgasr_priv.active = false; + vga_switcheroo_disable(); } mutex_unlock(&vgasr_mutex); } @@ -262,12 +351,13 @@ static int vga_switcheroo_show(struct seq_file *m, void *v) int i = 0; mutex_lock(&vgasr_mutex); list_for_each_entry(client, &vgasr_priv.clients, list) { - seq_printf(m, "%d:%s%s:%c:%s%s:%s\n", i, - client_id(client) == VGA_SWITCHEROO_DIS ? "DIS" : "IGD", + seq_printf(m, "%d:%s%s:%c:%s%s:%s:%s\n", i, + DIS_IGD_STR(client_id(client)), client_is_vga(client) ? "" : "-Audio", client->active ? '+' : ' ', client->driver_power_control ? "Dyn" : "", client->pwr_state ? "Pwr" : "Off", + client_id(client) == mux_state ? "mux" : "", pci_name(client->pdev)); i++; } @@ -304,6 +394,16 @@ static int vga_switchoff(struct vga_switcheroo_client *client) return 0; } +static int vga_switch_mux_to(int client_id) +{ + int ret = vgasr_priv.handler->switchto(client_id); + + if (ret == 0) + mux_state = client_id; + + return ret; +} + static void set_audio_state(int id, int state) { struct vga_switcheroo_client *client; @@ -353,7 +453,8 @@ static int vga_switchto_stage2(struct vga_switcheroo_client *new_client) console_unlock(); } - ret = vgasr_priv.handler->switchto(new_client->id); + ret = vga_switch_mux_to(new_client->id); + if (ret) return ret; @@ -468,7 +569,7 @@ vga_switcheroo_debugfs_write(struct file *filp, const char __user *ubuf, vgasr_priv.delayed_switch_active = false; if (just_mux) { - ret = vgasr_priv.handler->switchto(client_id); + ret = vga_switch_mux_to(client_id); goto out; } @@ -624,7 +725,7 @@ static int vga_switcheroo_runtime_suspend(struct device *dev) if (ret) return ret; if (vgasr_priv.handler->switchto) - vgasr_priv.handler->switchto(VGA_SWITCHEROO_IGD); + vga_switch_mux_to(VGA_SWITCHEROO_IGD); vga_switcheroo_power_switch(pdev, VGA_SWITCHEROO_OFF); return 0; } -- 2.1.1

1 0

[PATCH RFC v2] vga_switcheroo: restore mux and power states upon resume from hibernation.
by Stefan Demharter 26 Sep '14

26 Sep '14

Saves the current state of the mux and restores it upon resume from hibernation. This is needed for muxed systems because the state of the mux doesn't survive a hibernation-resume cycle. Furthermore also restores the power state of a GPU to OFF after resume from hibernation if it was OFF before hibernation. Signed-off-by: Stefan Demharter <stefan.demharter(a)gmx.net> --- drivers/gpu/vga/vga_switcheroo.c | 96 +++++++++++++++++++++++++++++++++++----- 1 file changed, 85 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/vga/vga_switcheroo.c b/drivers/gpu/vga/vga_switcheroo.c index ec0ae2d..85f41b9 100644 --- a/drivers/gpu/vga/vga_switcheroo.c +++ b/drivers/gpu/vga/vga_switcheroo.c @@ -28,6 +28,7 @@ #include <linux/console.h> #include <linux/vga_switcheroo.h> #include <linux/pm_runtime.h> +#include <linux/suspend.h> #include <linux/vgaarb.h> @@ -72,6 +73,64 @@ static struct vgasr_priv vgasr_priv = { .clients = LIST_HEAD_INIT(vgasr_priv.clients), }; +#define MUX_STATE_UNDEFINED (-1) +/* + * The next variable stores the state of the mux + * so it can be restored upon resume from hibernation. + */ +static int mux_state = MUX_STATE_UNDEFINED; + +static void vga_restore_mux(void) +{ + if (mux_state != MUX_STATE_UNDEFINED) { + int ret; + pr_info("vga_switcheroo: restoring mux state to %s\n", + mux_state == VGA_SWITCHEROO_IGD ? "IGD" : "DIS"); + + ret = vgasr_priv.handler->switchto(mux_state); + if (ret) + pr_warn("vga_switcheroo: mux switch failed (%d)\n", ret); + } +} + +static void vga_restore_power_state(void) +{ + struct vga_switcheroo_client *client; + + if (!vgasr_priv.handler->power_state) + return; + + list_for_each_entry(client, &vgasr_priv.clients, list) { + if (client_is_vga(client) && client->pwr_state == VGA_SWITCHEROO_OFF) { + pr_info("vga_switcheroo: setting power state of %s to OFF\n", + client->id == VGA_SWITCHEROO_IGD ? "IGD" : "DIS"); + vgasr_priv.handler->power_state(client->id, client->pwr_state); + } + } +} + +/* + * Handle resume from hibernation: + * Restore mux state and power states for switcheroo controlled GPUs. + */ +static int vga_resume(struct notifier_block *nb, unsigned long action, void *unused) +{ + switch (action) { + case PM_POST_HIBERNATION: + mutex_lock(&vgasr_mutex); + vga_restore_mux(); + vga_restore_power_state(); + mutex_unlock(&vgasr_mutex); + break; + } + return 0; +} + +static struct notifier_block vga_switcheroo_pm_nb = { + .notifier_call = vga_resume, + .priority = 0, +}; + static bool vga_switcheroo_ready(void) { /* we're ready if we get two clients + handler */ @@ -84,6 +143,8 @@ static void vga_switcheroo_enable(void) int ret; struct vga_switcheroo_client *client; + pr_info("vga_switcheroo: enabled\n"); + /* call the handler to init */ if (vgasr_priv.handler->init) vgasr_priv.handler->init(); @@ -99,6 +160,15 @@ static void vga_switcheroo_enable(void) } vga_switcheroo_debugfs_init(&vgasr_priv); vgasr_priv.active = true; + register_pm_notifier(&vga_switcheroo_pm_nb); +} + +void vga_switcheroo_disable(void) +{ + unregister_pm_notifier(&vga_switcheroo_pm_nb); + vgasr_priv.active = false; + vga_switcheroo_debugfs_fini(&vgasr_priv); + pr_info("vga_switcheroo: disabled\n"); } int vga_switcheroo_register_handler(struct vga_switcheroo_handler *handler) @@ -111,7 +181,6 @@ int vga_switcheroo_register_handler(struct vga_switcheroo_handler *handler) vgasr_priv.handler = handler; if (vga_switcheroo_ready()) { - printk(KERN_INFO "vga_switcheroo: enabled\n"); vga_switcheroo_enable(); } mutex_unlock(&vgasr_mutex); @@ -124,9 +193,7 @@ void vga_switcheroo_unregister_handler(void) mutex_lock(&vgasr_mutex); vgasr_priv.handler = NULL; if (vgasr_priv.active) { - pr_info("vga_switcheroo: disabled\n"); - vga_switcheroo_debugfs_fini(&vgasr_priv); - vgasr_priv.active = false; + vga_switcheroo_disable(); } mutex_unlock(&vgasr_mutex); } @@ -155,7 +222,6 @@ static int register_client(struct pci_dev *pdev, vgasr_priv.registered_clients++; if (vga_switcheroo_ready()) { - printk(KERN_INFO "vga_switcheroo: enabled\n"); vga_switcheroo_enable(); } mutex_unlock(&vgasr_mutex); @@ -235,9 +301,7 @@ void vga_switcheroo_unregister_client(struct pci_dev *pdev) kfree(client); } if (vgasr_priv.active && vgasr_priv.registered_clients < 2) { - printk(KERN_INFO "vga_switcheroo: disabled\n"); - vga_switcheroo_debugfs_fini(&vgasr_priv); - vgasr_priv.active = false; + vga_switcheroo_disable(); } mutex_unlock(&vgasr_mutex); } @@ -262,10 +326,11 @@ static int vga_switcheroo_show(struct seq_file *m, void *v) int i = 0; mutex_lock(&vgasr_mutex); list_for_each_entry(client, &vgasr_priv.clients, list) { - seq_printf(m, "%d:%s%s:%c:%s%s:%s\n", i, + seq_printf(m, "%d:%s%s:%c%s:%s%s:%s\n", i, client_id(client) == VGA_SWITCHEROO_DIS ? "DIS" : "IGD", client_is_vga(client) ? "" : "-Audio", client->active ? '+' : ' ', + client_id(client) == mux_state ? "mux" : "", client->driver_power_control ? "Dyn" : "", client->pwr_state ? "Pwr" : "Off", pci_name(client->pdev)); @@ -304,6 +369,15 @@ static int vga_switchoff(struct vga_switcheroo_client *client) return 0; } +static int vga_switchto(int client_id) +{ + int ret = vgasr_priv.handler->switchto(client_id); + + if (ret == 0) + mux_state = client_id; + return ret; +} + static void set_audio_state(int id, int state) { struct vga_switcheroo_client *client; @@ -353,7 +427,7 @@ static int vga_switchto_stage2(struct vga_switcheroo_client *new_client) console_unlock(); } - ret = vgasr_priv.handler->switchto(new_client->id); + ret = vga_switchto(new_client->id); if (ret) return ret; @@ -468,7 +542,7 @@ vga_switcheroo_debugfs_write(struct file *filp, const char __user *ubuf, vgasr_priv.delayed_switch_active = false; if (just_mux) { - ret = vgasr_priv.handler->switchto(client_id); + ret = vga_switchto(client_id); goto out; } -- 1.8.5.3

3 3

[Bug 69721] New: Can't reach maximum memory speed (or core speed) when using dpm=1 on r600g
by bugzilla-daemon＠freedesktop.org 26 Sep '14

26 Sep '14

https://bugs.freedesktop.org/show_bug.cgi?id=69721 Priority: medium Bug ID: 69721 Assignee: dri-devel(a)lists.freedesktop.org Summary: Can't reach maximum memory speed (or core speed) when using dpm=1 on r600g Severity: normal Classification: Unclassified OS: All Reporter: alexandre.f.demers(a)gmail.com Hardware: All Status: NEW Version: XOrg CVS Component: DRM/Radeon Product: DRI After fixing bug 68235, it appears the maximum memory speed (or core speed) can't always be reached. This situation depends on the voltage tweaking seen on some cards. As an exemple, XFX HD-695X-ZNDC has the following maximum values: max_sclk_vddc->80000, max_mclk_vddci->125000, max_mclk_vddc->125000. However, the maximum memory clock value should be 130000 and the core clock should be 83000. As Alex Deucher explained: "This patch set works around the issue by limiting the sclk and mclk to the highest levels listed in the clk/voltage dependency tables. I'll need to dig a bit more internally to try and figure out how to handle these clks properly." -- You are receiving this mail because: You are the assignee for the bug.

1 10

[PATCH] drm/edid: Add missing interlaced flag to 576i@100 modes.
by clinton.a.taylor＠intel.com 26 Sep '14

26 Sep '14

From: Clint Taylor <clinton.a.taylor(a)intel.com> CEA VICs 44 and 45 were missing DRM_MODE_FLAG_INTERLACE. Signed-off-by: Clint Taylor <clinton.a.taylor(a)intel.com> --- drivers/gpu/drm/drm_edid.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 1bdbfd0..3bf9991 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -841,13 +841,13 @@ static const struct drm_display_mode edid_cea_modes[] = { { DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732, 795, 864, 0, 576, 580, 586, 625, 0, DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC | - DRM_MODE_FLAG_DBLCLK), + DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK), .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, }, /* 45 - 720(1440)x576i@100Hz */ { DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732, 795, 864, 0, 576, 580, 586, 625, 0, DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC | - DRM_MODE_FLAG_DBLCLK), + DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK), .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, }, /* 46 - 1920x1080i@120Hz */ { DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008, -- 1.7.9.5

2 1

[PATCH v3 00/23] AMDKFD Kernel Driver
by Oded Gabbay 26 Sep '14

26 Sep '14

Hi, Here is the v3 patch set of amdkfd. This version contains changes and fixes to code, as agreed on during the review of the v2 patch set. The major changes are: - There are two new module parameters: # of processes and # of queues per process. The defaults, as agreed on in the v2 review, are 32 and 128 respectively. This sets the default amount of GART address space that amdkfd requires to 3.5MB (3MB for userspace queues mqds and 0.5MB for other stuff, such as mqd for kernel queue, hpd for pipelines, etc.) - All the GART address space usage of amdkfd is done inside a single contiguous buffer that is allocated from system memory, and pinned to the start of the GART during the startup of amdkfd (which is just after the startup of radeon). The management of this buffer is done by the radeon sa manager. This buffer is not evict-able. - Mapping of doorbells is initiated by the userspace lib (by mmap syscall), instead of initiating it from inside an ioctl (using vm_mmap). - Removed ioctls for exclusive access to performance counters - Added documentation about the QCM (Queue Control Management), apertures and interfaces between amdkfd and radeon. Two important notes: - The topology patch has not been changed. Look at http://lists.freedesktop.org/archives/dri-devel/2014-July/065042.html for my response. I also put my answer as an explanation in the commit msg of the patch. - There are still some minor code style issues I need to fix. I didn't want to delay v3 any further but I will publish either v4 with those fixes, or just relevant patches if the whole patch set will be merged. For people who like to review using git, the v3 patch set is located at: http://cgit.freedesktop.org/~gabbayo/linux/log/?h=kfd-next-3.17-v3 In addition, I would like to announce that we have uploaded the userspace lib that accompanies amdkfd. That lib is called "libhsakmt" and you can view it at: http://cgit.freedesktop.org/~gabbayo/libhsakmt Alexey Skidanov (1): amdkfd: Implement the Get Process Aperture IOCTL Andrew Lewycky (3): amdkfd: Add basic modules to amdkfd amdkfd: Add interrupt handling module amdkfd: Implement the Set Memory Policy IOCTL Ben Goz (8): amdkfd: Add queue module amdkfd: Add mqd_manager module amdkfd: Add kernel queue module amdkfd: Add module parameter of scheduling policy amdkfd: Add packet manager module amdkfd: Add process queue manager module amdkfd: Add device queue manager module amdkfd: Implement the create/destroy/update queue IOCTLs Evgeny Pinchuk (2): amdkfd: Add topology module to amdkfd amdkfd: Implement the Get Clock Counters IOCTL Oded Gabbay (9): drm/radeon: reduce number of free VMIDs and pipes in KV drm/radeon/cik: Don't touch int of pipes 1-7 drm/radeon: Report doorbell configuration to amdkfd drm/radeon: adding synchronization for GRBM GFX drm/radeon: Add radeon <--> amdkfd interface Update MAINTAINERS and CREDITS files with amdkfd info amdkfd: Add IOCTL set definitions of amdkfd amdkfd: Add amdkfd skeleton driver amdkfd: Add binding/unbinding calls to amd_iommu driver CREDITS | 7 + MAINTAINERS | 10 + drivers/gpu/drm/radeon/Kconfig | 2 + drivers/gpu/drm/radeon/Makefile | 3 + drivers/gpu/drm/radeon/amdkfd/Kconfig | 10 + drivers/gpu/drm/radeon/amdkfd/Makefile | 14 + drivers/gpu/drm/radeon/amdkfd/cik_regs.h | 220 ++++ drivers/gpu/drm/radeon/amdkfd/kfd_aperture.c | 350 ++++++ drivers/gpu/drm/radeon/amdkfd/kfd_chardev.c | 511 +++++++++ drivers/gpu/drm/radeon/amdkfd/kfd_crat.h | 294 +++++ drivers/gpu/drm/radeon/amdkfd/kfd_device.c | 300 +++++ .../drm/radeon/amdkfd/kfd_device_queue_manager.c | 989 ++++++++++++++++ .../drm/radeon/amdkfd/kfd_device_queue_manager.h | 144 +++ drivers/gpu/drm/radeon/amdkfd/kfd_doorbell.c | 236 ++++ drivers/gpu/drm/radeon/amdkfd/kfd_interrupt.c | 161 +++ drivers/gpu/drm/radeon/amdkfd/kfd_kernel_queue.c | 330 ++++++ drivers/gpu/drm/radeon/amdkfd/kfd_kernel_queue.h | 66 ++ drivers/gpu/drm/radeon/amdkfd/kfd_module.c | 147 +++ drivers/gpu/drm/radeon/amdkfd/kfd_mqd_manager.c | 305 +++++ drivers/gpu/drm/radeon/amdkfd/kfd_mqd_manager.h | 88 ++ drivers/gpu/drm/radeon/amdkfd/kfd_packet_manager.c | 495 ++++++++ drivers/gpu/drm/radeon/amdkfd/kfd_pasid.c | 95 ++ drivers/gpu/drm/radeon/amdkfd/kfd_pm4_headers.h | 682 +++++++++++ drivers/gpu/drm/radeon/amdkfd/kfd_pm4_opcodes.h | 107 ++ drivers/gpu/drm/radeon/amdkfd/kfd_priv.h | 560 +++++++++ drivers/gpu/drm/radeon/amdkfd/kfd_process.c | 347 ++++++ .../drm/radeon/amdkfd/kfd_process_queue_manager.c | 346 ++++++ drivers/gpu/drm/radeon/amdkfd/kfd_queue.c | 85 ++ drivers/gpu/drm/radeon/amdkfd/kfd_topology.c | 1207 ++++++++++++++++++++ drivers/gpu/drm/radeon/amdkfd/kfd_topology.h | 168 +++ drivers/gpu/drm/radeon/cik.c | 154 +-- drivers/gpu/drm/radeon/cik_reg.h | 65 ++ drivers/gpu/drm/radeon/cikd.h | 53 +- drivers/gpu/drm/radeon/radeon.h | 10 + drivers/gpu/drm/radeon/radeon_device.c | 32 + drivers/gpu/drm/radeon/radeon_drv.c | 5 + drivers/gpu/drm/radeon/radeon_kfd.c | 525 +++++++++ drivers/gpu/drm/radeon/radeon_kfd.h | 177 +++ drivers/gpu/drm/radeon/radeon_kms.c | 7 + include/uapi/linux/kfd_ioctl.h | 126 ++ 40 files changed, 9338 insertions(+), 95 deletions(-) create mode 100644 drivers/gpu/drm/radeon/amdkfd/Kconfig create mode 100644 drivers/gpu/drm/radeon/amdkfd/Makefile create mode 100644 drivers/gpu/drm/radeon/amdkfd/cik_regs.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_aperture.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_chardev.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_crat.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_device.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_device_queue_manager.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_device_queue_manager.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_doorbell.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_interrupt.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_kernel_queue.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_kernel_queue.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_module.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_mqd_manager.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_mqd_manager.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_packet_manager.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_pasid.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_pm4_headers.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_pm4_opcodes.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_priv.h create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_process.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_process_queue_manager.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_queue.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_topology.c create mode 100644 drivers/gpu/drm/radeon/amdkfd/kfd_topology.h create mode 100644 drivers/gpu/drm/radeon/radeon_kfd.c create mode 100644 drivers/gpu/drm/radeon/radeon_kfd.h create mode 100644 include/uapi/linux/kfd_ioctl.h -- 1.9.1

4 28

[PATCH 0/5] drm/nouveau: platform devices and GK20A probing
by Alexandre Courbot 26 Sep '14

26 Sep '14

This patch series is the final (?) step towards the initial support of GK20A, allowing it to be probed and used (currently at a very slow speed, and for offscreen rendering only) on the Jetson TK1 and Venice 2 boards. The main piece if the first patch which adds platform devices probing support to Nouveau. There are probably lots of things that need to be discussed about it, e.g.: * The way the DRM device is created, especially with respect to the ongoing changes to the DRM framework, * The fact that the same drm_driver instance is used for the PCI and platform drivers, * Whether we should have only one platform driver capable of probing all platform devices, or one driver per GPU (in this case, where should all these drivers reside?) So there are still some rough edges, but we are getting there. :) The first patch should go through the Nouveau tree, while the 4 others are rather intended for Tegra. Alexandre Courbot (3): drm/nouveau: support for probing platform devices ARM: tegra: of: add GK20A device tree binding ARM: tegra: jetson-tk1: enable GK20A GPU Thierry Reding (2): ARM: tegra: add GK20A GPU to Tegra124 DT ARM: tegra: venice2: enable GK20A GPU .../devicetree/bindings/gpu/nvidia,gk20a.txt | 45 +++++ arch/arm/boot/dts/tegra124-jetson-tk1.dts | 8 +- arch/arm/boot/dts/tegra124-venice2.dts | 8 +- arch/arm/boot/dts/tegra124.dtsi | 15 ++ drivers/gpu/drm/nouveau/Kconfig | 8 + drivers/gpu/drm/nouveau/Makefile | 3 + drivers/gpu/drm/nouveau/nouveau_drm.c | 33 ++-- drivers/gpu/drm/nouveau/nouveau_drm.h | 21 +++ drivers/gpu/drm/nouveau/nouveau_platform.c | 191 +++++++++++++++++++++ 9 files changed, 315 insertions(+), 17 deletions(-) create mode 100644 Documentation/devicetree/bindings/gpu/nvidia,gk20a.txt create mode 100644 drivers/gpu/drm/nouveau/nouveau_platform.c -- 1.9.2

6 18

[PATCH v7 0/3] Add drm driver for Rockchip Socs
by Mark yao 26 Sep '14

26 Sep '14

This a series of patches is a DRM Driver for Rockchip Socs, add support for vop devices. Future patches will add additional encoders/connectors, such as eDP, HDMI. The basic "crtc" for rockchip is a "VOP" - Video Output Processor. the vop devices found on Rockchip rk3288 Soc, rk3288 soc have two similar Vop devices. Vop devices support iommu mapping, we use dma-mapping API with ARM_DMA_USE_IOMMU. Changes in v2: - add DRM master device node to list all display nodes that comprise the graphics subsystem. - use the component framework to defer main drm driver probe until all VOP devices have been probed. - use dma-mapping API with ARM_DMA_USE_IOMMU, create dma mapping by master device and each vop device can shared the drm dma mapping. - use drm_crtc_init_with_planes and drm_universal_plane_init. - remove unnecessary middle layers. - add cursor set, move funcs to rockchip drm crtc. - add vop reset. Changes in v3: - change "crtc->fb" to "crtc->primary-fb" Adviced by Daniel Vetter - init cursor plane with universal api, remove unnecessary cursor set,move Changes in v4: Adviced by David Herrmann - remove drm_platform_*() usage, use register drm device directly. Adviced by Rob Clark - remove special mmap ioctl, do userspace mmap with normal mmap() or mmap offset Changes in v5: Adviced by Arnd Bergmann - doing DMA start with a 32-bit masks with dma_mask and dma_coherent_mark - fix some incorrect dependencies. Adviced by Boris BREZILLON - fix some mistake and bugs. Adviced by Daniel Vetter - drop all special ioctl and use generic kms ioctl instead. Adviced by Rob Clark - use unlocked api for drm_fb_helper_restore_fbdev_mode. - remove unused rockchip_gem_prime_import_sg_table. Changes in v6: - set gem buffer pitch 64 bytes align, needed by mali gpu. Adviced by Daniel Kurtz - fix some mistake, bugs, remove unused define, more better code style etc. - use clk_prepare()/unprepare() at probe()/remove() and clk_enable()/disable() at runtime instead of clk_prepare_enable(). - provide a help function from vop for encoder to do mode config, instead of using drm_diaplay_mode private method. - change vop mode_set timing to make it more safely. Changes in v7: - fix memory leakage problem. Mark yao (3): drm/rockchip: Add basic drm driver dt-bindings: video: Add for rockchip display subsytem dt-bindings: video: Add documentation for rockchip vop .../devicetree/bindings/video/rockchip-drm.txt | 19 + .../devicetree/bindings/video/rockchip-vop.txt | 58 + drivers/gpu/drm/Kconfig | 2 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/rockchip/Kconfig | 17 + drivers/gpu/drm/rockchip/Makefile | 8 + drivers/gpu/drm/rockchip/rockchip_drm_drv.c | 516 +++++++ drivers/gpu/drm/rockchip/rockchip_drm_drv.h | 61 + drivers/gpu/drm/rockchip/rockchip_drm_fb.c | 201 +++ drivers/gpu/drm/rockchip/rockchip_drm_fb.h | 28 + drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c | 230 ++++ drivers/gpu/drm/rockchip/rockchip_drm_fbdev.h | 20 + drivers/gpu/drm/rockchip/rockchip_drm_gem.c | 345 +++++ drivers/gpu/drm/rockchip/rockchip_drm_gem.h | 55 + drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 1422 ++++++++++++++++++++ drivers/gpu/drm/rockchip/rockchip_drm_vop.h | 196 +++ 16 files changed, 3179 insertions(+) -- 1.7.9.5

1 3

[Bug 71859] New: texelFetch segfault in libLLVM-3.3.so (on Cayman)
by bugzilla-daemon＠freedesktop.org 26 Sep '14

26 Sep '14

https://bugs.freedesktop.org/show_bug.cgi?id=71859 Priority: medium Bug ID: 71859 Assignee: dri-devel(a)lists.freedesktop.org Summary: texelFetch segfault in libLLVM-3.3.so (on Cayman) Severity: normal Classification: Unclassified OS: All Reporter: alexandre.f.demers(a)gmail.com Hardware: Other Status: NEW Version: git Component: Drivers/Gallium/r600 Product: Mesa Nov 20 22:55:37 Xander kernel: texelFetch[18811]: segfault at 0 ip 00007f41626876ab sp 00007fff5b0e7400 error 6 in libLLVM-3.3.so[7f4162326000+158d000] Nov 20 22:55:37 Xander systemd-coredump[18832]: Process 18811 (texelFetch) dumped core. Nov 20 22:55:40 Xander systemd-coredump[19099]: Process 19087 (texelFetch) dumped core. Nov 20 22:55:45 Xander systemd-coredump[19393]: Process 19375 (texelFetch) dumped core. Nov 20 22:55:46 Xander systemd-coredump[19453]: Process 19434 (texelFetch) dumped core. Nov 20 22:55:55 Xander systemd-coredump[20162]: Process 20145 (texelFetch) dumped core. Nov 20 22:56:02 Xander systemd-coredump[20619]: Process 20605 (texelFetch) dumped core. Nov 20 22:56:11 Xander systemd-coredump[21328]: Process 21307 (arb_texture_mul) dumped core. Nov 20 22:56:11 Xander systemd-coredump[21329]: Process 21309 (arb_texture_mul) dumped core. Nov 20 22:56:11 Xander systemd-coredump[21332]: Process 21318 (arb_texture_mul) dumped core. Nov 20 22:56:11 Xander systemd-coredump[21352]: Process 21333 (arb_texture_mul) dumped core. Nov 20 22:56:26 Xander systemd-coredump[22297]: Process 22281 (texelFetch) dumped core. Nov 20 22:56:35 Xander systemd-coredump[22874]: Process 22860 (texelFetch) dumped core. Nov 20 22:56:38 Xander systemd-coredump[23056]: Process 23038 (ext_transform_f) dumped core. Nov 20 22:56:41 Xander systemd-coredump[23235]: Process 23221 (texelFetch) dumped core. Nov 20 22:56:41 Xander systemd-coredump[23257]: Process 23242 (texelFetch) dumped core. Nov 20 22:56:45 Xander systemd-coredump[23484]: Process 23468 (texelFetch) dumped core. Nov 20 22:56:54 Xander systemd-coredump[23978]: Process 23960 (glx-create-cont) dumped core. Nov 20 22:57:17 Xander systemd-coredump[25666]: Process 25643 (texelFetch) dumped core. Nov 20 22:57:17 Xander systemd-coredump[25721]: Process 25704 (ext_transform_f) dumped core. -- You are receiving this mail because: You are the assignee for the bug.

1 12

[PATCH v5 0/3] Add drm driver for Rockchip Socs
by Mark yao 26 Sep '14

26 Sep '14

This a series of patches is a DRM Driver for Rockchip Socs, add support for vop devices. Future patches will add additional encoders/connectors, such as eDP, HDMI. The basic "crtc" for rockchip is a "VOP" - Video Output Processor. the vop devices found on Rockchip rk3288 Soc, rk3288 soc have two similar Vop devices. Vop devices support iommu mapping, we use dma-mapping API with ARM_DMA_USE_IOMMU. Changes in v2: - add DRM master device node to list all display nodes that comprise the graphics subsystem. - use the component framework to defer main drm driver probe until all VOP devices have been probed. - use dma-mapping API with ARM_DMA_USE_IOMMU, create dma mapping by master device and each vop device can shared the drm dma mapping. - use drm_crtc_init_with_planes and drm_universal_plane_init. - remove unnecessary middle layers. - add cursor set, move funcs to rockchip drm crtc. - add vop reset. Changes in v3: - change "crtc->fb" to "crtc->primary-fb" Adviced by Daniel Vetter - init cursor plane with universal api, remove unnecessary cursor set,move Changes in v4: Adviced by David Herrmann - remove drm_platform_*() usage, use register drm device directly. Adviced by Rob Clark - remove special mmap ioctl, do userspace mmap with normal mmap() or mmap offset Changes in v5: Adviced by Arnd Bergmann - doing DMA start with a 32-bit masks with dma_mask and dma_coherent_mark - fix some incorrect dependencies. Adviced by Boris BREZILLON - fix some mistake and bugs. Adviced by Daniel Vetter - drop all special ioctl and use generic kms ioctl instead. Adviced by Rob Clark - use unlocked api for drm_fb_helper_restore_fbdev_mode. - remove unused rockchip_gem_prime_import_sg_table. Mark yao (3): drm/rockchip: Add basic drm driver dt-bindings: video: Add for rockchip display subsytem dt-bindings: video: Add documentation for rockchip vop .../devicetree/bindings/video/rockchip-drm.txt | 19 + .../devicetree/bindings/video/rockchip-vop.txt | 58 + drivers/gpu/drm/Kconfig | 2 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/rockchip/Kconfig | 17 + drivers/gpu/drm/rockchip/Makefile | 8 + drivers/gpu/drm/rockchip/rockchip_drm_drv.c | 509 ++++++++ drivers/gpu/drm/rockchip/rockchip_drm_drv.h | 120 ++ drivers/gpu/drm/rockchip/rockchip_drm_fb.c | 201 +++ drivers/gpu/drm/rockchip/rockchip_drm_fb.h | 28 + drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c | 230 ++++ drivers/gpu/drm/rockchip/rockchip_drm_fbdev.h | 20 + drivers/gpu/drm/rockchip/rockchip_drm_gem.c | 341 +++++ drivers/gpu/drm/rockchip/rockchip_drm_gem.h | 55 + drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 1373 ++++++++++++++++++++ drivers/gpu/drm/rockchip/rockchip_drm_vop.h | 187 +++ 16 files changed, 3169 insertions(+) -- 1.7.9.5

3 8

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel September 2014