enh: when reading CPUID is unavailable (VM?), fallback to cpuinfo where applicable

built from commit 954eb13468
 dated 2026-04-06 18:58:36 +0200
 by Stéphane Lesimple (speed47_github@speed47.net)

 cap_* variable <= cpuinfo flag

cap_ibrs              <= ibrs
cap_ibpb              <= ibpb
cap_stibp             <= stibp
cap_ssbd              <= ssbd / virt_ssbd
cap_l1df              <= flush_l1d
cap_md_clear          <= md_clear
cap_arch_capabilities <= arch_capabilities

Should fix #288

.github/workflows/expected_cve_count

@@ -1 +1 @@
-26
+31

README.md

@@ -22,12 +22,17 @@ CVE | Name | Aliases
 [CVE-2019-11091](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2019-11091) | Microarchitectural Data Sampling Uncacheable Memory | MDSUM, RIDL
 [CVE-2019-11135](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2019-11135) | TSX Asynchronous Abort | TAA, ZombieLoad V2
 [CVE-2020-0543](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-0543) | Special Register Buffer Data Sampling | SRBDS, CROSSTalk
+[CVE-2022-21123](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-21123) | Shared Buffers Data Read | SBDR, MMIO Stale Data
+[CVE-2022-21125](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-21125) | Shared Buffers Data Sampling | SBDS, MMIO Stale Data
+[CVE-2022-21166](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-21166) | Device Register Partial Write | DRPW, MMIO Stale Data
 [CVE-2022-29900](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-29900) | Arbitrary Speculative Code Execution with Return Instructions | Retbleed (AMD)
 [CVE-2022-29901](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-29901) | Arbitrary Speculative Code Execution with Return Instructions | Retbleed (Intel), RSBA
 [CVE-2022-40982](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2022-40982) | Gather Data Sampling | Downfall, GDS
 [CVE-2023-20569](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2023-20569) | Return Address Security | Inception, SRSO
+[CVE-2023-20588](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2023-20588) | AMD Division by Zero Speculative Data Leak | DIV0
 [CVE-2023-20593](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2023-20593) | Cross-Process Information Leak | Zenbleed
 [CVE-2023-23583](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2023-23583) | Redundant Prefix Issue | Reptar
+[CVE-2023-28746](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2023-28746) | Register File Data Sampling | RFDS
 [CVE-2024-28956](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2024-28956) | Indirect Target Selection | ITS
 [CVE-2024-36350](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2024-36350) | Transient Scheduler Attack, Store Queue | TSA-SQ
 [CVE-2024-36357](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2024-36357) | Transient Scheduler Attack, L1 | TSA-L1
@@ -56,12 +61,17 @@ CVE-2018-12207 (iTLB Multihit, No eXcuses) | ✅ | ✅ | ☠️ | ✅ | Hypervis
 CVE-2019-11091 (MDSUM, RIDL) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
 CVE-2019-11135 (TAA, ZombieLoad V2) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
 CVE-2020-0543 (SRBDS, CROSSTalk) | 💥 (2) | 💥 (2) | 💥 (2) | 💥 (2) | Microcode + kernel update
+CVE-2022-21123 (SBDR, MMIO Stale Data) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
+CVE-2022-21125 (SBDS, MMIO Stale Data) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
+CVE-2022-21166 (DRPW, MMIO Stale Data) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
 CVE-2022-29900 (Retbleed AMD) | 💥 | ✅ | 💥 | ✅ | Kernel update (+ microcode for IBPB)
 CVE-2022-29901 (Retbleed Intel, RSBA) | 💥 | ✅ | 💥 | ✅ | Microcode + kernel update (eIBRS or IBRS)
 CVE-2022-40982 (Downfall, GDS) | 💥 | 💥 | 💥 | 💥 | Microcode update (or disable AVX)
 CVE-2023-20569 (Inception, SRSO) | 💥 | ✅ | 💥 | ✅ | Microcode + kernel update
+CVE-2023-20588 (DIV0) | 💥 | 💥 (1) | 💥 | 💥 (1) | Kernel update (+ disable SMT)
 CVE-2023-20593 (Zenbleed) | 💥 | 💥 | 💥 | 💥 | Microcode update (or kernel workaround)
 CVE-2023-23583 (Reptar) | ☠️ | ☠️ | ☠️ | ☠️ | Microcode update
+CVE-2023-28746 (RFDS) | 💥 | ✅ | 💥 | ✅ | Microcode + kernel update
 CVE-2024-28956 (ITS) | 💥 | ✅ | 💥 (4) | ✅ | Microcode + kernel update
 CVE-2024-36350 (TSA-SQ) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
 CVE-2024-36357 (TSA-L1) | 💥 | 💥 (1) | 💥 | 💥 (1) | Microcode + kernel update
@@ -141,6 +151,10 @@ On CPUs with Intel TSX, a transactional abort can leave data from the line fill
 
 Certain special CPU instructions (RDRAND, RDSEED, EGETKEY) read data through a shared staging buffer that is accessible across all cores via speculative execution. An attacker running code on any core can observe the output of these instructions from a victim on a different core, including extracting cryptographic keys from SGX enclaves (a complete ECDSA key was demonstrated). This is notable as one of the first cross-core speculative execution attacks. Mitigation requires a microcode update that serializes access to the staging buffer, plus a kernel update to manage the mitigation. Performance impact is low, mainly affecting workloads that heavily use RDRAND/RDSEED.
 
+**CVE-2022-21123, CVE-2022-21125, CVE-2022-21166 — Processor MMIO Stale Data (SBDR, SBDS, DRPW)**
+
+A class of MMIO (Memory-Mapped I/O) vulnerabilities where stale data from CPU internal fill buffers can be inferred through side-channel attacks during MMIO operations. Three sub-vulnerabilities are covered: Shared Buffers Data Read (SBDR, CVE-2022-21123), Shared Buffers Data Sampling (SBDS, CVE-2022-21125), and Device Register Partial Write (DRPW, CVE-2022-21166). Affected Intel CPUs include Haswell through Rocket Lake server and client processors, plus Tremont Atom cores. Mitigation requires a microcode update providing the FB_CLEAR capability (VERW instruction clears fill buffers) plus a kernel update (Linux 5.19+) that invokes VERW at kernel/user transitions and VM entry/exit. When SMT is enabled, sibling threads can still exploit the vulnerability even with mitigations active. Performance impact is low, as the VERW mechanism is shared with the existing MDS mitigation.
+
 **CVE-2022-29900 — Arbitrary Speculative Code Execution with Return Instructions (Retbleed AMD)**
 
 On AMD processors from families 0x15 through 0x17 (Bulldozer through Zen 2) and Hygon family 0x18, an attacker can exploit return instructions to redirect speculative execution and leak kernel memory, bypassing retpoline mitigations that were effective against Spectre V2. Unlike Spectre V2 which targets indirect jumps and calls, Retbleed specifically targets return instructions, which were previously considered safe. Mitigation requires a kernel update providing either the untrained return thunk (safe RET) or IBPB-on-entry mechanism, plus a microcode update providing IBPB support on Zen 1/2. On Zen 1/2, SMT should be disabled for full protection when using IBPB-based mitigation. Performance impact is medium.
@@ -157,6 +171,10 @@ The AVX GATHER instructions can leak data from previously used vector registers
 
 On AMD Zen 1 through Zen 4 processors, an attacker can manipulate the return address predictor to redirect speculative execution on return instructions, leaking kernel memory. Mitigation requires both a kernel update (providing SRSO safe-return sequences or IBPB-on-entry) and a microcode update (providing SBPB on Zen 3/4, or IBPB support on Zen 1/2 — which additionally requires SMT to be disabled). Performance impact ranges from low to significant depending on the chosen mitigation and CPU generation.
 
+**CVE-2023-20588 — AMD Division by Zero Speculative Data Leak (DIV0)**
+
+On AMD Zen 1 processors, a #DE (divide-by-zero) exception can leave stale quotient data from a previous division in the divider unit, observable by a subsequent division via speculative side channels. This can leak data across any privilege boundary, including between SMT sibling threads sharing the same physical core. Mitigation requires a kernel update (Linux 6.5+) that adds a dummy division (`amd_clear_divider()`) on every exit to userspace and before VMRUN, preventing stale data from persisting. No microcode update is needed. Disabling SMT provides additional protection because the kernel mitigation does not cover cross-SMT-thread leaks. Performance impact is negligible.
+
 **CVE-2023-20593 — Cross-Process Information Leak (Zenbleed)**
 
 A bug in AMD Zen 2 processors causes the VZEROUPPER instruction to incorrectly zero register files during speculative execution, leaving stale data from other processes observable in vector registers. This can leak data across any privilege boundary, including from the kernel and other processes, at rates up to 30 KB/s per core. Mitigation is available either through a microcode update that fixes the bug, or through a kernel workaround that sets the FP_BACKUP_FIX bit (bit 9) in the DE_CFG MSR, disabling the faulty optimization. Either approach alone is sufficient. Performance impact is negligible.
@@ -165,6 +183,10 @@ A bug in AMD Zen 2 processors causes the VZEROUPPER instruction to incorrectly z
 
 A bug in Intel processors causes unexpected behavior when executing instructions with specific redundant REX prefixes. Depending on the circumstances, this can result in a system crash (MCE), unpredictable behavior, or potentially privilege escalation. Any software running on an affected CPU can trigger the bug. Mitigation requires a microcode update. Performance impact is low.
 
+**CVE-2023-28746 — Register File Data Sampling (RFDS)**
+
+On certain Intel Atom and hybrid processors (Goldmont, Goldmont Plus, Tremont, Gracemont, and the Atom cores of Alder Lake and Raptor Lake), the register file can retain stale data from previous operations that is accessible via speculative execution, allowing an attacker to infer data across privilege boundaries. Mitigation requires both a microcode update (providing the RFDS_CLEAR capability) and a kernel update (CONFIG_MITIGATION_RFDS, Linux 6.9+) that uses the VERW instruction to clear the register file on privilege transitions. CPUs with the RFDS_NO capability bit are not affected. Performance impact is low.
+
 **CVE-2024-28956 — Indirect Target Selection (ITS)**
 
 On certain Intel processors (Skylake-X stepping 6+, Kaby Lake, Comet Lake, Ice Lake, Tiger Lake, Rocket Lake), an attacker can train the indirect branch predictor to speculatively execute a targeted gadget in the kernel, bypassing eIBRS protections. The Branch Target Buffer (BTB) uses only partial address bits to index indirect branch targets, allowing user-space code to influence kernel-space speculative execution. Some affected CPUs (Ice Lake, Tiger Lake, Rocket Lake) are only vulnerable to native user-to-kernel attacks, not guest-to-host (VMX) attacks. Mitigation requires both a microcode update (IPU 2025.1 / microcode-20250512+, which fixes IBPB to fully flush indirect branch predictions) and a kernel update (CONFIG_MITIGATION_ITS, Linux 6.15+) that aligns branch/return thunks or uses RSB stuffing. Performance impact is low.