<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: 12pt; color: rgb(0, 0, 0);">
Hi Dan Carpenter,</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Thank you for your review, there is indeed a problem with NULL Pointers, I will correct this problem immediately</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Regards,<br>
Sun,Ce</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> Dan Carpenter <dan.carpenter@linaro.org><br>
<b>Sent:</b> Wednesday, April 9, 2025 4:46 PM<br>
<b>To:</b> Sun, Ce(Overlord) <Ce.Sun@amd.com><br>
<b>Cc:</b> amd-gfx@lists.freedesktop.org <amd-gfx@lists.freedesktop.org>; SHANMUGAM, SRINIVASAN <SRINIVASAN.SHANMUGAM@amd.com><br>
<b>Subject:</b> [bug report] drm/amdgpu: Multi-GPU DPC recovery support</font>
<div> </div>
</div>
<div class="BodyFragment"><font size="2"><span style="font-size:11pt;">
<div class="PlainText">Hello Ce Sun,<br>
<br>
Commit 8ba904f54148 ("drm/amdgpu: Multi-GPU DPC recovery support")<br>
from Mar 21, 2025 (linux-next), leads to the following Smatch static<br>
checker warning:<br>
<br>
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:6820 amdgpu_pci_slot_reset()<br>
warn: iterator used outside loop: 'tmp_adev'<br>
<br>
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c<br>
6753 pci_ers_result_t amdgpu_pci_slot_reset(struct pci_dev *pdev)<br>
6754 {<br>
6755 struct drm_device *dev = pci_get_drvdata(pdev);<br>
6756 struct amdgpu_device *adev = drm_to_adev(dev);<br>
6757 struct amdgpu_reset_context reset_context;<br>
6758 struct amdgpu_device *tmp_adev = NULL;<br>
6759 struct amdgpu_hive_info *hive = NULL;<br>
6760 struct list_head device_list;<br>
6761 int r = 0, i;<br>
6762 u32 memsize;<br>
6763 <br>
6764 /* PCI error slot reset should be skipped During RAS recovery */<br>
6765 if ((amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3) ||<br>
6766 amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4)) &&<br>
6767 amdgpu_ras_in_recovery(adev))<br>
6768 return PCI_ERS_RESULT_RECOVERED;<br>
6769 <br>
6770 dev_info(adev->dev, "PCI error: slot reset callback!!\n");<br>
6771 <br>
6772 memset(&reset_context, 0, sizeof(reset_context));<br>
6773 <br>
6774 /* wait for asic to come out of reset */<br>
6775 msleep(700);<br>
6776 <br>
6777 /* Restore PCI confspace */<br>
6778 amdgpu_device_load_pci_state(pdev);<br>
6779 <br>
6780 /* confirm ASIC came out of reset */<br>
6781 for (i = 0; i < adev->usec_timeout; i++) {<br>
6782 memsize = amdgpu_asic_get_config_memsize(adev);<br>
6783 <br>
6784 if (memsize != 0xffffffff)<br>
6785 break;<br>
6786 udelay(1);<br>
6787 }<br>
6788 if (memsize == 0xffffffff) {<br>
6789 r = -ETIME;<br>
6790 goto out;<br>
6791 }<br>
6792 <br>
6793 reset_context.method = AMD_RESET_METHOD_NONE;<br>
6794 reset_context.reset_req_dev = adev;<br>
6795 set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);<br>
6796 set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags);<br>
6797 INIT_LIST_HEAD(&device_list);<br>
6798 <br>
6799 hive = amdgpu_get_xgmi_hive(adev);<br>
6800 if (hive) {<br>
6801 mutex_lock(&hive->hive_lock);<br>
6802 reset_context.hive = hive;<br>
6803 list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) {<br>
6804 tmp_adev->pcie_reset_ctx.in_link_reset = true;<br>
6805 list_add_tail(&tmp_adev->reset_list, &device_list);<br>
6806 }<br>
<br>
tmp_adev is an invalid non-NULL pointer.<br>
<br>
6807 } else {<br>
6808 set_bit(AMDGPU_SKIP_HW_RESET, &reset_context.flags);<br>
6809 list_add_tail(&adev->reset_list, &device_list);<br>
6810 }<br>
6811 <br>
6812 r = amdgpu_device_asic_reset(adev, &device_list, &reset_context);<br>
6813 out:<br>
6814 if (!r) {<br>
6815 if (amdgpu_device_cache_pci_state(adev->pdev))<br>
6816 pci_restore_state(adev->pdev);<br>
6817 dev_info(adev->dev, "PCIe error recovery succeeded\n");<br>
6818 } else {<br>
6819 dev_err(adev->dev, "PCIe error recovery failed, err:%d\n", r);<br>
--> 6820 if (tmp_adev) {<br>
<br>
This looks like it might have been intentional but it would be more<br>
readable to check if (hive) {<br>
<br>
6821 list_for_each_entry(tmp_adev, &device_list, reset_list)<br>
6822 amdgpu_device_unset_mp1_state(tmp_adev);<br>
6823 amdgpu_device_unlock_reset_domain(adev->reset_domain);<br>
6824 }<br>
6825 }<br>
6826 <br>
6827 if (hive) {<br>
6828 mutex_unlock(&hive->hive_lock);<br>
6829 amdgpu_put_xgmi_hive(hive);<br>
6830 }<br>
6831 <br>
6832 return r ? PCI_ERS_RESULT_DISCONNECT : PCI_ERS_RESULT_RECOVERED;<br>
6833 }<br>
<br>
regards,<br>
dan carpenter<br>
</div>
</span></font></div>
</div>
</body>
</html>