From ea5ce46fb4e5f285f049ac7a02ce42a97aeb028f Mon Sep 17 00:00:00 2001
From: Joseph Greathouse <Joseph.Greathouse@amd.com>
Date: Fri, 29 Oct 2021 09:24:05 -0500
Subject: [PATCH] Add better PCI ID backup in rocm_agent_enumerator

The PCI ID backup method in rocm_agent_enumerator, where the
tool uses lspci to find all AMD GPU devices in the system and
manaully match them to gfx version, is extremely outdated. The
PCI ID list did not include anything after Vega 10, and the
actual call to lspci no longer returned anything due to some
missing conversions.

The patch adds all GPUs that might be needed by ROCr up through
Navy Flounder. The PCI ID to gfx matching pulls from the amdgpu
driver and libhsakmt.

Change-Id: I58b77bb6aa631f575352fc444d2542f265909706
---
 rocm_agent_enumerator | 55 ++++++++++++++++++++++++++++++++++++-------
 1 file changed, 47 insertions(+), 8 deletions(-)

diff --git a/rocm_agent_enumerator b/rocm_agent_enumerator
index 5295873097..3d8028a9e5 100755
--- a/rocm_agent_enumerator
+++ b/rocm_agent_enumerator
@@ -10,7 +10,9 @@ CWD = os.path.dirname(os.path.realpath(__file__))
 
 ISA_TO_ID = {
   # Kaveri - Temporary
-  "gfx700" : [0x130f],
+  "gfx700" : [0x1304, 0x1305, 0x1306, 0x1307, 0x1309, 0x130a, 0x130b, 0x130c,
+              0x130d, 0x130e, 0x130f, 0x1310, 0x1311, 0x1312, 0x1313, 0x1315,
+              0x1316, 0x1317, 0x1318, 0x131b, 0x131c, 0x131d],
   # Hawaii
   "gfx701" : [0x67a0, 0x67a1, 0x67a2, 0x67a8, 0x67a9, 0x67aa, 0x67b0, 0x67b1,
               0x67b8, 0x67b9, 0x67ba, 0x67be],
@@ -23,15 +25,52 @@ ISA_TO_ID = {
   "gfx803" : [0x7300, 0x730f,
   # Polaris10
               0x67c0, 0x67c1, 0x67c2, 0x67c4, 0x67c7, 0x67c8, 0x67c9, 0x67ca,
-              0x67cc, 0x67cf,
+              0x67cc, 0x67cf, 0x6fdf,
   # Polaris11
               0x67d0, 0x67df, 0x67e0, 0x67e1, 0x67e3, 0x67e7, 0x67e8, 0x67e9,
               0x67eb, 0x67ef, 0x67ff,
   # Polaris12
-              0x6980, 0x6981, 0x6985, 0x6986, 0x6987, 0x6995, 0x6997, 0x699f],
+              0x6980, 0x6981, 0x6985, 0x6986, 0x6987, 0x6995, 0x6997, 0x699f,
+  # VegaM
+              0x694c, 0x694e, 0x694f],
   # Vega10
-  "gfx900" : [0x6860, 0x6861, 0x6862, 0x6863, 0x6864, 0x6867, 0x6868, 0x686c,
-              0x687f]
+  "gfx900" : [0x6860, 0x6861, 0x6862, 0x6863, 0x6864, 0x6867, 0x6868, 0x6869,
+              0x6869, 0x686a, 0x686b, 0x686c, 0x686d, 0x686e, 0x686f, 0x687f],
+  # Raven
+  "gfx902" : [0x15dd, 0x15d8],
+  # Vega12
+  "gfx904" : [0x69a0, 0x69a1, 0x69a2, 0x69a3, 0x69af],
+  # Vega20
+  "gfx906" : [0x66a0, 0x66a1, 0x66a2, 0x66a3, 0x66a4, 0x66a7, 0x66af],
+  # Arcturus
+  "gfx908" : [0x738c, 0x7388, 0x738e, 0x7390],
+  # Aldebaran
+  "gfx90a" : [0x7408, 0x740c, 0x740f, 0x7410],
+  # Renoir
+  "gfx90c" : [0x15e7, 0x1636, 0x1638, 0x164c],
+  # Navi10
+  "gfx1010" : [0x7310, 0x7312, 0x7318, 0x7319, 0x731a, 0x731b, 0x731e, 0x731f],
+  # Navi12
+  "gfx1011" : [0x7360, 0x7362],
+  # Navi14
+  "gfx1012" : [0x7340, 0x7341, 0x7347, 0x734f],
+  # Cyan_Skillfish
+  "gfx1013" : [0x13f9, 0x13fa, 0x13fb, 0x13fc, 0x13f3],
+  # Sienna_Cichlid
+  "gfx1030" : [0x73a0, 0x73a1, 0x73a2, 0x73a3, 0x73a5, 0x73a8, 0x73a9, 0x73ab,
+               0x73ac, 0x73ad, 0x73ae, 0x73af, 0x73bf],
+  # Navy_Flounder
+  "gfx1031" : [0x73c0, 0x73c1, 0x73c3, 0x73da, 0x73db, 0x73dc, 0x73dd, 0x73de,
+               0x73df],
+  # Dimgray_Cavefish
+  "gfx1032" : [0x73e0, 0x73e1, 0x73e2, 0x73e3, 0x73e8, 0x73e9, 0x73ea, 0x73eb,
+               0x73ec, 0x73ed, 0x73ef, 0x73ff],
+  # Van Gogh
+  "gfx1033" : [0x163f],
+  # Beige_Goby
+  "gfx1034" : [0x7420, 0x7421, 0x7422, 0x7423, 0x743f],
+  # Yellow_Carp
+  "gfx1035" : [0x164d, 0x1681]
 }
 
 def staticVars(**kwargs):
@@ -112,7 +151,7 @@ def readFromLSPCI():
 
   try:
     # run lspci
-    lspci_output = subprocess.Popen(["/usr/bin/lspci", "-n", "-d", "1002:"], stdout=subprocess.PIPE).communicate()[0].split('\n')
+    lspci_output = subprocess.Popen(["/usr/bin/lspci", "-n", "-d", "1002:"], stdout=subprocess.PIPE).communicate()[0].decode("utf-8").split('\n')
   except:
     lspci_output = []
 
@@ -160,8 +199,8 @@ def main():
     if len(target_list) == 0:
       target_list = readFromLSPCI()
 
-  # workaround to cope with existing rocm_agent_enumerator behavior where gfx000
-  # would always be returned
+    # workaround to cope with existing rocm_agent_enumerator behavior where gfx000
+    # would always be returned
     print("gfx000")
 
   for gfx in target_list: