<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<p style="font-family:Calibri;font-size:10pt;color:#0000FF;margin:5pt;font-style:normal;font-weight:normal;text-decoration:none;" align="Left">
[AMD Official Use Only - AMD Internal Distribution Only]<br>
</p>
<br>
<div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
Signed-off-by: Tony Yi <Tony.Yi@amd.com></div>
<div id="appendonsend"></div>
<hr style="display:inline-block;width:98%" tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" style="font-size:11pt" color="#000000"><b>From:</b> Zhang, Hawking <Hawking.Zhang@amd.com><br>
<b>Sent:</b> Wednesday, April 2, 2025 12:23 AM<br>
<b>To:</b> Skvortsov, Victor <Victor.Skvortsov@amd.com>; amd-gfx@lists.freedesktop.org <amd-gfx@lists.freedesktop.org><br>
<b>Cc:</b> Luo, Zhigang <Zhigang.Luo@amd.com>; Zhou1, Tao <Tao.Zhou1@amd.com>; Zhao, Victor <Victor.Zhao@amd.com>; Yi, Tony <Tony.Yi@amd.com><br>
<b>Subject:</b> RE: [PATCH] drm/amdgpu: Fix CPER error handling on VFs</font>
<div> </div>
</div>
<div class="BodyFragment"><font face="Times New Roman" size="3"><span style="font-size:12pt;"><a name="BM_BEGIN"></a>
<div><font size="2"><span style="font-size:11pt;">[AMD Official Use Only - AMD Internal Distribution Only]<br>
<br>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com><br>
<br>
Regards,<br>
Hawking<br>
-----Original Message-----<br>
From: Skvortsov, Victor <Victor.Skvortsov@amd.com><br>
Sent: Wednesday, April 2, 2025 04:44<br>
To: amd-gfx@lists.freedesktop.org<br>
Cc: Luo, Zhigang <Zhigang.Luo@amd.com>; Zhang, Hawking <Hawking.Zhang@amd.com>; Zhou1, Tao <Tao.Zhou1@amd.com>; Zhao, Victor <Victor.Zhao@amd.com>; Yi, Tony <Tony.Yi@amd.com>; Skvortsov, Victor <Victor.Skvortsov@amd.com><br>
Subject: [PATCH] drm/amdgpu: Fix CPER error handling on VFs<br>
<br>
From: Tony Yi <Tony.Yi@amd.com><br>
<br>
CPER read will loop infinitely if an error is encountered and the more bit is set. Add error checks to break upon failure.<br>
<br>
Suggested-by: Tony Yi <Tony.Yi@amd.com><br>
Signed-off-by: Victor Skvortsov <Victor.Skvortsov@amd.com><br>
---<br>
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 8 +++++---<br>
 1 file changed, 5 insertions(+), 3 deletions(-)<br>
<br>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c<br>
index 0bb8cbe0dcc0..8d2da3a27440 100644<br>
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c<br>
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c<br>
@@ -1378,14 +1378,16 @@ amdgpu_virt_write_cpers_to_ring(struct amdgpu_device *adev,<br>
        used_size = host_telemetry->header.used_size;<br>
<br>
        if (used_size > (AMD_SRIOV_RAS_TELEMETRY_SIZE_KB << 10))<br>
-               return 0;<br>
+               return -EINVAL;<br>
<br>
        cper_dump = kmemdup(&host_telemetry->body.cper_dump, used_size, GFP_KERNEL);<br>
        if (!cper_dump)<br>
                return -ENOMEM;<br>
<br>
-       if (checksum != amd_sriov_msg_checksum(cper_dump, used_size, 0, 0))<br>
+       if (checksum != amd_sriov_msg_checksum(cper_dump, used_size, 0, 0)) {<br>
+               ret = -EINVAL;<br>
                goto out;<br>
+       }<br>
<br>
        *more = cper_dump->more;<br>
<br>
@@ -1434,7 +1436,7 @@ static int amdgpu_virt_req_ras_cper_dump_internal(struct amdgpu_device *adev)<br>
                                adev, virt->fw_reserve.ras_telemetry, &more);<br>
                else<br>
                        ret = 0;<br>
-       } while (more);<br>
+       } while (more && !ret);<br>
<br>
        return ret;<br>
 }<br>
--<br>
2.34.1<br>
<br>
</span></font></div>
</span></font></div>
</div>
</body>
</html>