<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<meta name="Generator" content="Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:#0563C1;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:#954F72;
text-decoration:underline;}
p.MsoPlainText, li.MsoPlainText, div.MsoPlainText
{mso-style-priority:99;
mso-style-link:"Plain Text Char";
margin:0in;
margin-bottom:.0001pt;
font-size:14.0pt;
font-family:"Calibri",sans-serif;}
span.PlainTextChar
{mso-style-name:"Plain Text Char";
mso-style-priority:99;
mso-style-link:"Plain Text";
font-family:"Calibri",sans-serif;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang="EN-US" link="#0563C1" vlink="#954F72">
<div class="WordSection1">
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">-----Original Message-----<br>
From: Grodzovsky, Andrey <Andrey.Grodzovsky@amd.com> <br>
Sent: Wednesday, November 27, 2019 11:46 PM<br>
To: Ma, Le <Le.Ma@amd.com>; amd-gfx@lists.freedesktop.org<br>
Cc: Chen, Guchun <Guchun.Chen@amd.com>; Zhou1, Tao <Tao.Zhou1@amd.com>; Deucher, Alexander <Alexander.Deucher@amd.com>; Li, Dennis <Dennis.Li@amd.com>; Zhang, Hawking <Hawking.Zhang@amd.com><br>
Subject: Re: [PATCH 07/10] drm/amdgpu: add concurrent baco reset support for XGMI</p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">On 11/27/19 4:15 AM, Le Ma wrote:<o:p></o:p></p>
<p class="MsoPlainText">> Currently each XGMI node reset wq does not run in parrallel because
<o:p></o:p></p>
<p class="MsoPlainText">> same work item bound to same cpu runs in sequence. So change to bound
<o:p></o:p></p>
<p class="MsoPlainText">> the xgmi_reset_work item to different cpus.<o:p></o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">It's not the same work item, see more bellow<o:p></o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> XGMI requires all nodes enter into baco within very close proximity
<o:p></o:p></p>
<p class="MsoPlainText">> before any node exit baco. So schedule the xgmi_reset_work wq twice
<o:p></o:p></p>
<p class="MsoPlainText">> for enter/exit baco respectively.<o:p></o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> The default reset code path and methods do not change for vega20 production:<o:p></o:p></p>
<p class="MsoPlainText">> - baco reset without xgmi/ras<o:p></o:p></p>
<p class="MsoPlainText">> - psp reset with xgmi/ras<o:p></o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> To enable baco for XGMI/RAS case, both 2 conditions below are needed:<o:p></o:p></p>
<p class="MsoPlainText">> - amdgpu_ras_enable=2<o:p></o:p></p>
<p class="MsoPlainText">> - baco-supported smu firmware<o:p></o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> The case that PSP reset and baco reset coexist within an XGMI hive is
<o:p></o:p></p>
<p class="MsoPlainText">> not in the consideration.<o:p></o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> Change-Id: I9c08cf90134f940b42e20d2129ff87fba761c532<o:p></o:p></p>
<p class="MsoPlainText">> Signed-off-by: Le Ma <<a href="mailto:le.ma@amd.com"><span style="color:windowtext;text-decoration:none">le.ma@amd.com</span></a>><o:p></o:p></p>
<p class="MsoPlainText">> ---<o:p></o:p></p>
<p class="MsoPlainText">> drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 +<o:p></o:p></p>
<p class="MsoPlainText">> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 78 ++++++++++++++++++++++++++----<o:p></o:p></p>
<p class="MsoPlainText">> 2 files changed, 70 insertions(+), 10 deletions(-)<o:p></o:p></p>
<p class="MsoPlainText">><o:p> </o:p></p>
<p class="MsoPlainText">> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h <o:p></o:p></p>
<p class="MsoPlainText">> b/drivers/gpu/drm/amd/amdgpu/amdgpu.h<o:p></o:p></p>
<p class="MsoPlainText">> index d120fe5..08929e6 100644<o:p></o:p></p>
<p class="MsoPlainText">> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h<o:p></o:p></p>
<p class="MsoPlainText">> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h<o:p></o:p></p>
<p class="MsoPlainText">> @@ -998,6 +998,8 @@ struct amdgpu_device {<o:p></o:p></p>
<p class="MsoPlainText">> int pstate;<o:p></o:p></p>
<p class="MsoPlainText">> /* enable runtime pm on the device */<o:p></o:p></p>
<p class="MsoPlainText">> bool runpm;<o:p></o:p></p>
<p class="MsoPlainText">> +<o:p></o:p></p>
<p class="MsoPlainText">> + bool in_baco;<o:p></o:p></p>
<p class="MsoPlainText">> };<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> static inline struct amdgpu_device *amdgpu_ttm_adev(struct
<o:p></o:p></p>
<p class="MsoPlainText">> ttm_bo_device *bdev) diff --git <o:p></o:p></p>
<p class="MsoPlainText">> a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c <o:p></o:p></p>
<p class="MsoPlainText">> b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c<o:p></o:p></p>
<p class="MsoPlainText">> index bd387bb..71abfe9 100644<o:p></o:p></p>
<p class="MsoPlainText">> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c<o:p></o:p></p>
<p class="MsoPlainText">> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c<o:p></o:p></p>
<p class="MsoPlainText">> @@ -2654,7 +2654,13 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)<o:p></o:p></p>
<p class="MsoPlainText">> struct amdgpu_device *adev =<o:p></o:p></p>
<p class="MsoPlainText">> container_of(__work, struct amdgpu_device, xgmi_reset_work);<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> - adev->asic_reset_res = amdgpu_asic_reset(adev);<o:p></o:p></p>
<p class="MsoPlainText">> + if (amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)<o:p></o:p></p>
<p class="MsoPlainText">> + adev->asic_reset_res = (adev->in_baco == false) ?<o:p></o:p></p>
<p class="MsoPlainText">> + amdgpu_device_baco_enter(adev->ddev) :<o:p></o:p></p>
<p class="MsoPlainText">> + amdgpu_device_baco_exit(adev->ddev);<o:p></o:p></p>
<p class="MsoPlainText">> + else<o:p></o:p></p>
<p class="MsoPlainText">> + adev->asic_reset_res = amdgpu_asic_reset(adev);<o:p></o:p></p>
<p class="MsoPlainText">> +<o:p></o:p></p>
<p class="MsoPlainText">> if (adev->asic_reset_res)<o:p></o:p></p>
<p class="MsoPlainText">> DRM_WARN("ASIC reset failed with error, %d for drm dev, %s",<o:p></o:p></p>
<p class="MsoPlainText">> adev->asic_reset_res, adev->ddev->unique); @@ -3796,6 +3802,7 @@
<o:p></o:p></p>
<p class="MsoPlainText">> static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,<o:p></o:p></p>
<p class="MsoPlainText">> struct amdgpu_device *tmp_adev = NULL;<o:p></o:p></p>
<p class="MsoPlainText">> bool need_full_reset = *need_full_reset_arg, vram_lost = false;<o:p></o:p></p>
<p class="MsoPlainText">> int r = 0;<o:p></o:p></p>
<p class="MsoPlainText">> + int cpu = smp_processor_id();<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> /*<o:p></o:p></p>
<p class="MsoPlainText">> * ASIC reset has to be done on all HGMI hive nodes ASAP @@
<o:p></o:p></p>
<p class="MsoPlainText">> -3803,21 +3810,24 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,<o:p></o:p></p>
<p class="MsoPlainText">> */<o:p></o:p></p>
<p class="MsoPlainText">> if (need_full_reset) {<o:p></o:p></p>
<p class="MsoPlainText">> list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) {<o:p></o:p></p>
<p class="MsoPlainText">> - /* For XGMI run all resets in parallel to speed up the process */<o:p></o:p></p>
<p class="MsoPlainText">> + /*<o:p></o:p></p>
<p class="MsoPlainText">> + * For XGMI run all resets in parallel to speed up the<o:p></o:p></p>
<p class="MsoPlainText">> + * process by scheduling the highpri wq on different<o:p></o:p></p>
<p class="MsoPlainText">> + * cpus. For XGMI with baco reset, all nodes must enter<o:p></o:p></p>
<p class="MsoPlainText">> + * baco within close proximity before anyone exit.<o:p></o:p></p>
<p class="MsoPlainText">> + */<o:p></o:p></p>
<p class="MsoPlainText">> if (tmp_adev->gmc.xgmi.num_physical_nodes > 1) {<o:p></o:p></p>
<p class="MsoPlainText">> - if (!queue_work(system_highpri_wq, &tmp_adev->xgmi_reset_work))<o:p></o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">Note that tmp_adev->xgmi_reset_work (the work item) is per device in XGMI hive and not the same work item. So I don't see why you need to explicitly queue them on different CPUs, they should run in parallel already.<o:p></o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">Andrey<o:p></o:p></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">[Le]: It’s also beyond my understanding that the 2 node reset work items scheduled to same cpu does not run in parallel. But from
the experiment result in my side, the 2nd work item always run after 1st work item finished. Based on this result, I changed to queue them on different CPUs to make sure more XGMI nodes case to run in parallel, because baco requires all nodes enter baco within
very close proximity. <o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"><o:p> </o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">The experiment code is as following for your reference. When card0 worker running, card1 worker is not observed to run.<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"><o:p> </o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+atomic_t card0_in_baco = ATOMIC_INIT(0);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+atomic_t card1_in_baco = ATOMIC_INIT(0);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">{<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> struct amdgpu_device *adev =<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> container_of(__work, struct amdgpu_device, xgmi_reset_work);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"><o:p> </o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ printk("lema1: card 0x%x goes into reset wq\n", adev->pdev->bus->number);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ if (adev->pdev->bus->number == 0x7) {<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ atomic_set(&card1_in_baco, 1);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ printk("lema1: card1 in baco from card1 view\n");<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ }<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> if (amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> adev->asic_reset_res = (adev->in_baco == false) ?<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> amdgpu_device_baco_enter(adev->ddev) :<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">@@ -2664,6 +2673,23 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> if (adev->asic_reset_res)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> DRM_WARN("ASIC reset failed with error, %d for drm dev, %s",<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%"> adev->asic_reset_res, adev->ddev->unique);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ if (adev->pdev->bus->number == 0x4) {<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ atomic_set(&card0_in_baco, 1);<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ printk("lema1: card0 in baco from card0 view\n");<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ while (true)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ if (!!atomic_read(&card1_in_baco))<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ break;<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ printk("lema1: card1 in baco from card0 view\n");<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ }<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ if (adev->pdev->bus->number == 0x7) {<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ while (true)<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ if (!!atomic_read(&card0_in_baco))<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ break;<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ printk("lema1: card0 in baco from card1 view\n");<o:p></o:p></span></p>
<p class="MsoPlainText"><span style="color:#203864;mso-style-textfill-fill-color:#203864;mso-style-textfill-fill-alpha:100.0%">+ }<o:p></o:p></span></p>
<p class="MsoPlainText"><o:p> </o:p></p>
<p class="MsoPlainText">> + if (!queue_work_on(cpu, system_highpri_wq,<o:p></o:p></p>
<p class="MsoPlainText">> + &tmp_adev->xgmi_reset_work))<o:p></o:p></p>
<p class="MsoPlainText">> r = -EALREADY;<o:p></o:p></p>
<p class="MsoPlainText">> + cpu = cpumask_next(cpu, cpu_online_mask);<o:p></o:p></p>
<p class="MsoPlainText">> } else<o:p></o:p></p>
<p class="MsoPlainText">> r = amdgpu_asic_reset(tmp_adev);<o:p></o:p></p>
<p class="MsoPlainText">> -<o:p></o:p></p>
<p class="MsoPlainText">> - if (r) {<o:p></o:p></p>
<p class="MsoPlainText">> - DRM_ERROR("ASIC reset failed with error, %d for drm dev, %s",<o:p></o:p></p>
<p class="MsoPlainText">> - r, tmp_adev->ddev->unique);<o:p></o:p></p>
<p class="MsoPlainText">> + if (r)<o:p></o:p></p>
<p class="MsoPlainText">> break;<o:p></o:p></p>
<p class="MsoPlainText">> - }<o:p></o:p></p>
<p class="MsoPlainText">> }<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> - /* For XGMI wait for all PSP resets to complete before proceed */<o:p></o:p></p>
<p class="MsoPlainText">> + /* For XGMI wait for all work to complete before proceed */<o:p></o:p></p>
<p class="MsoPlainText">> if (!r) {<o:p></o:p></p>
<p class="MsoPlainText">> list_for_each_entry(tmp_adev, device_list_handle,<o:p></o:p></p>
<p class="MsoPlainText">> gmc.xgmi.head) {<o:p></o:p></p>
<p class="MsoPlainText">> @@ -3826,11 +3836,59 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,<o:p></o:p></p>
<p class="MsoPlainText">> r = tmp_adev->asic_reset_res;<o:p></o:p></p>
<p class="MsoPlainText">> if (r)<o:p></o:p></p>
<p class="MsoPlainText">> break;<o:p></o:p></p>
<p class="MsoPlainText">> + if(AMD_RESET_METHOD_BACO ==<o:p></o:p></p>
<p class="MsoPlainText">> + amdgpu_asic_reset_method(tmp_adev))<o:p></o:p></p>
<p class="MsoPlainText">> + tmp_adev->in_baco = true;<o:p></o:p></p>
<p class="MsoPlainText">> }<o:p></o:p></p>
<p class="MsoPlainText">> }<o:p></o:p></p>
<p class="MsoPlainText">> }<o:p></o:p></p>
<p class="MsoPlainText">> - }<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> + /*<o:p></o:p></p>
<p class="MsoPlainText">> + * For XGMI with baco reset, need exit baco phase by scheduling<o:p></o:p></p>
<p class="MsoPlainText">> + * xgmi_reset_work one more time. PSP reset skips this phase.<o:p></o:p></p>
<p class="MsoPlainText">> + * Not assume the situation that PSP reset and baco reset<o:p></o:p></p>
<p class="MsoPlainText">> + * coexist within an XGMI hive.<o:p></o:p></p>
<p class="MsoPlainText">> + */<o:p></o:p></p>
<p class="MsoPlainText">> +<o:p></o:p></p>
<p class="MsoPlainText">> + if (!r) {<o:p></o:p></p>
<p class="MsoPlainText">> + cpu = smp_processor_id();<o:p></o:p></p>
<p class="MsoPlainText">> + list_for_each_entry(tmp_adev, device_list_handle,<o:p></o:p></p>
<p class="MsoPlainText">> + gmc.xgmi.head) {<o:p></o:p></p>
<p class="MsoPlainText">> + if (tmp_adev->gmc.xgmi.num_physical_nodes > 1<o:p></o:p></p>
<p class="MsoPlainText">> + && AMD_RESET_METHOD_BACO ==<o:p></o:p></p>
<p class="MsoPlainText">> + amdgpu_asic_reset_method(tmp_adev)) {<o:p></o:p></p>
<p class="MsoPlainText">> + if (!queue_work_on(cpu,<o:p></o:p></p>
<p class="MsoPlainText">> + system_highpri_wq,<o:p></o:p></p>
<p class="MsoPlainText">> + &tmp_adev->xgmi_reset_work))<o:p></o:p></p>
<p class="MsoPlainText">> + r = -EALREADY;<o:p></o:p></p>
<p class="MsoPlainText">> + if (r)<o:p></o:p></p>
<p class="MsoPlainText">> + break;<o:p></o:p></p>
<p class="MsoPlainText">> + cpu = cpumask_next(cpu, cpu_online_mask);<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> +<o:p></o:p></p>
<p class="MsoPlainText">> + if (!r) {<o:p></o:p></p>
<p class="MsoPlainText">> + list_for_each_entry(tmp_adev, device_list_handle,<o:p></o:p></p>
<p class="MsoPlainText">> + gmc.xgmi.head) {<o:p></o:p></p>
<p class="MsoPlainText">> + if (tmp_adev->gmc.xgmi.num_physical_nodes > 1<o:p></o:p></p>
<p class="MsoPlainText">> + && AMD_RESET_METHOD_BACO ==<o:p></o:p></p>
<p class="MsoPlainText">> + amdgpu_asic_reset_method(tmp_adev)) {<o:p></o:p></p>
<p class="MsoPlainText">> + flush_work(&tmp_adev->xgmi_reset_work);<o:p></o:p></p>
<p class="MsoPlainText">> + r = tmp_adev->asic_reset_res;<o:p></o:p></p>
<p class="MsoPlainText">> + if (r)<o:p></o:p></p>
<p class="MsoPlainText">> + break;<o:p></o:p></p>
<p class="MsoPlainText">> + tmp_adev->in_baco = false;<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> +<o:p></o:p></p>
<p class="MsoPlainText">> + if (r) {<o:p></o:p></p>
<p class="MsoPlainText">> + DRM_ERROR("ASIC reset failed with error, %d for drm dev, %s",<o:p></o:p></p>
<p class="MsoPlainText">> + r, tmp_adev->ddev->unique);<o:p></o:p></p>
<p class="MsoPlainText">> + goto end;<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> + }<o:p></o:p></p>
<p class="MsoPlainText">> <o:p></o:p></p>
<p class="MsoPlainText">> list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) {<o:p></o:p></p>
<p class="MsoPlainText">> if (need_full_reset) {<o:p></o:p></p>
</div>
</body>
</html>