Gentoo Archives: gentoo-user

From: Mick <michaelkintzios@×××××.com>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] Re: AMD microcode updates - where are they?!
Date: Sat, 13 Jul 2019 10:02:18
Message-Id: 6206681.QGZE5AxUjX@localhost
In Reply to: Re: [gentoo-user] Re: AMD microcode updates - where are they?! by Adam Carter
1 Thank you both, for your replies.
2
3 On Saturday, 13 July 2019 01:56:30 BST Adam Carter wrote:
4 > > > $ dmesg | grep -i micro
5 > > > [ 0.622441] [drm] Loading ARUBA Microcode
6 > > > [ 5.763242] [drm] Loading hainan Microcode
7 > > > [ 6.653025] microcode: CPU0: patch_level=0x06001119
8 > > > [ 6.657962] microcode: CPU1: patch_level=0x06001119
9 > > > [ 6.658890] microcode: CPU2: patch_level=0x06001119
10 > > > [ 6.659881] microcode: CPU3: patch_level=0x06001119
11 > > > [ 6.661136] microcode: Microcode Update Driver: v2.2.
12 > >
13 > > I have a similar experience:
14 > >
15 > > [ 0.659996] microcode: CPU0: patch_level=0x010000c8
16 > > [ 0.660001] microcode: CPU1: patch_level=0x010000c8
17 > > [ 0.660006] microcode: CPU2: patch_level=0x010000c8
18 > > [ 0.660011] microcode: CPU3: patch_level=0x010000c8
19 > > [ 0.660029] microcode: Microcode Update Driver: v2.2.
20 > > [ 7.853509] [drm] Loading RS780 Microcode
21 > >
22 > > I have a 10h generation processor, and I also build in microcode_amd.bin
23 > > with the kernel.
24
25 I had not until now built in 'amd-ucode/microcode_amd.bin', only 'amd-ucode/
26 microcode_amd_fam15h.bin', because this laptop has a 15h family CPU:
27
28 # lscpu
29 Architecture: x86_64
30 CPU op-mode(s): 32-bit, 64-bit
31 Byte Order: Little Endian
32 Address sizes: 48 bits physical, 48 bits virtual
33 CPU(s): 4
34 On-line CPU(s) list: 0-3
35 Thread(s) per core: 2
36 Core(s) per socket: 2
37 Socket(s): 1
38 NUMA node(s): 1
39 Vendor ID: AuthenticAMD
40 CPU family: 21
41 Model: 19
42 Model name: AMD A10-5750M APU with Radeon(tm) HD Graphics
43 Stepping: 1
44 CPU MHz: 1330.218
45 CPU max MHz: 2500.0000
46 CPU min MHz: 1400.0000
47 BogoMIPS: 4990.70
48 Virtualization: AMD-V
49 L1d cache: 16K
50 L1i cache: 64K
51 L2 cache: 2048K
52 NUMA node0 CPU(s): 0-3
53 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
54 cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb
55 rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf
56 pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 popcnt aes xsave avx f16c
57 lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch
58 osvw ibs xop skinit wdt lwp fma4 tce nodeid_msr tbm topoext perfctr_core
59 perfctr_nb cpb hw_pstate ssbd vmmcall bmi1 arat npt lbrv svm_lock nrip_save
60 tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
61
62 $ dmesg | grep family:
63 [ 0.291910] smpboot: CPU0: AMD A10-5750M APU with Radeon(tm) HD Graphics
64 (family: 0x15, model: 0x13, stepping: 0x1)
65
66
67 > Piledriver gets the early message barcelona/fam10h doesnt;
68 >
69 > # dmesg | grep microc
70 > [ 1.663099] microcode: microcode updated early to new
71 > patch_level=0x06000852
72 > [ 1.664161] microcode: CPU0: patch_level=0x06000852
73 > [ 1.665147] microcode: CPU1: patch_level=0x06000852
74 > [ 1.666135] microcode: CPU2: patch_level=0x06000852
75 > [ 1.667119] microcode: CPU3: patch_level=0x06000852
76 > [ 1.668034] microcode: CPU4: patch_level=0x06000852
77 > [ 1.668955] microcode: CPU5: patch_level=0x06000852
78 > [ 1.670060] microcode: CPU6: patch_level=0x06000852
79 > [ 1.670985] microcode: CPU7: patch_level=0x06000852
80 > [ 1.672012] microcode: Microcode Update Driver: v2.2.
81
82 OK, mine is also a Piledriver (mobile) CPU according to these tables:
83
84 https://en.wikichip.org/wiki/amd/a10
85
86 However, I don't see any early microcode being loaded. :-/
87
88 I added 'amd-ucode/microcode_amd.bin' in the kernel, just in case it was
89 needed and rebooted, but still no difference.
90
91 > # dmesg | grep microc
92 > [ 1.700378] microcode: CPU0: patch_level=0x010000c8
93 > [ 1.700435] microcode: CPU1: patch_level=0x010000c8
94 > [ 1.700488] microcode: CPU2: patch_level=0x010000c8
95 > [ 1.700543] microcode: CPU3: patch_level=0x010000c8
96 > [ 1.700684] microcode: Microcode Update Driver: v2.2.
97 >
98 > microcode_amd.bin hasn't changed since at least January 2018, so maybe
99 > there hasnt been any updates for the recent CPU vulnerabilities.
100 >
101 > Assuming the numbering is sequential its odd that the APU is at 0x06001119
102 > but the latest from linux-firmware is only 0x010000c8. Are you sure the APU
103 > is not fam16h ?
104
105 Yes, positive. As I've shown above the laptop has an A10 fam15h mobile
106 Piledriver processor.
107
108 The desktop has an A10 Steamroller, Kaveri APU:
109
110 $ dmesg | grep family:
111 [ 0.269754] smpboot: CPU0: AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G
112 (family: 0x15, model: 0x30, stepping: 0x1)
113
114 It has the 'amd-ucode/microcode_amd_fam15h.bin' built in the kernel and it
115 also shows no early microcode being loaded:
116
117 $ dmesg | grep micro
118 [ 1.578553] microcode: CPU0: patch_level=0x06003106
119 [ 1.579338] microcode: CPU1: patch_level=0x06003106
120 [ 1.580943] microcode: CPU2: patch_level=0x06003106
121 [ 1.581729] microcode: CPU3: patch_level=0x06003106
122 [ 1.582608] microcode: Microcode Update Driver: v2.2.
123
124 Notice the patch number is slightly higher than the laptop's, which can be
125 explained by the fact the desktop's Steamroller was launched in Jan 2014,
126 while the laptop's Piledriver was launched in March 2013.
127
128 Anyway, this does not explain why your Piledriver at least is loading early
129 the microcode, but mine isn't. :-/
130
131 --
132 Regards,
133
134 Mick

Attachments

File name MIME type
signature.asc application/pgp-signature