GPU (Windows): use a faster implementation

This commit is contained in:
李通洲
2023-03-25 13:04:44 +08:00
parent 945b5f1b15
commit a524302ea7
4 changed files with 63 additions and 177 deletions
+3 -3
View File
@@ -36,11 +36,11 @@ static void pciDetectVendorName(FFGPUResult* gpu, PCIData* pci, struct pci_dev*
pci->ffpci_lookup_name(pci->access, gpu->vendor.chars, (int) gpu->vendor.allocated, PCI_LOOKUP_VENDOR, device->vendor_id);
ffStrbufRecalculateLength(&gpu->vendor);
if(ffStrbufFirstIndexS(&gpu->vendor, "AMD") < gpu->vendor.length || ffStrbufFirstIndexS(&gpu->vendor, "ATI") < gpu->vendor.length)
if(ffStrbufContainS(&gpu->vendor, "AMD") || ffStrbufContainS(&gpu->vendor, "ATI"))
ffStrbufSetS(&gpu->vendor, FF_GPU_VENDOR_NAME_AMD);
else if(ffStrbufFirstIndexS(&gpu->vendor, "Intel") < gpu->vendor.length)
else if(ffStrbufContainS(&gpu->vendor, "Intel"))
ffStrbufSetS(&gpu->vendor, FF_GPU_VENDOR_NAME_INTEL);
else if(ffStrbufFirstIndexS(&gpu->vendor, "NVIDIA") < gpu->vendor.length)
else if(ffStrbufContainS(&gpu->vendor, "NVIDIA"))
ffStrbufSetS(&gpu->vendor, FF_GPU_VENDOR_NAME_NVIDIA);
}