El-errata: ELSA-2022-9273 Important: Oracle Linux 8 Unbreakable Enterprise kernel security update
Oracle Linux Security Advisory ELSA-2022-9273
http://linux.oracle.com/errata/ELSA-2022-9273.html
The following updated rpms for Oracle Linux 8 have been uploaded to the Unbreakable Linux Network:
x86_64:
kernel-uek-5.4.17-2136.306.1.3.el8uek.x86_64.rpm
kernel-uek-debug-5.4.17-2136.306.1.3.el8uek.x86_64.rpm
kernel-uek-debug-devel-5.4.17-2136.306.1.3.el8uek.x86_64.rpm
kernel-uek-devel-5.4.17-2136.306.1.3.el8uek.x86_64.rpm
kernel-uek-doc-5.4.17-2136.306.1.3.el8uek.noarch.rpm
SRPMS:
http://oss.oracle.com/ol8/SRPMS-updates/kernel-uek-5.4.17-2136.306.1.3.el8uek.src.rpm
Related CVEs:
CVE-2020-36516
CVE-2021-22600
CVE-2021-26341
CVE-2021-26401
CVE-2022-0617
CVE-2022-1016
CVE-2022-1158
CVE-2022-22942
CVE-2022-23960
CVE-2022-24448
CVE-2022-26966
Description of changes:
[5.4.17-2136.306.1.3.el8uek]
- KVM: x86/mmu: do compare-and-exchange of gPTE via the user address (Paolo Bonzini) [Orabug: 34053807] {CVE-2022-1158}
[5.4.17-2136.306.1.2.el8uek]
- Revert "rds/ib: recover rds connection from stuck rx path" (Rohit Nair) [Orabug: 34045203]
[5.4.17-2136.306.1.1.el8uek]
- netfilter: nf_tables: initialize registers in nft_do_chain() (Pablo Neira Ayuso) [Orabug: 34035701] {CVE-2022-1016}
- exec, elf: ignore invalid note data (Anthony Yznaga) [Orabug: 34035682]
[5.4.17-2136.306.1.el8uek]
- sr9700: sanity check for packet length (Brian Maly) [Orabug: 33962705] {CVE-2022-26966}
- net/packet: rx_owner_map depends on pg_vec (Willem de Bruijn) [Orabug: 33835787] {CVE-2021-22600}
- KVM: SVM: Don't flush cache if hardware enforces cache coherency across encryption domains (Krish Sadhukhan) [Orabug: 33921125]
- x86/mm/pat: Don't flush cache if hardware enforces cache coherency across encryption domnains (Krish Sadhukhan) [Orabug: 33921125]
- rds/ib: Resize CQ if send-/recv-ring-size are changed (Hans Westgaard Ry) [Orabug: 33940519]
- NFSv4: Handle case where the lookup of a directory fails (Trond Myklebust) [Orabug: 33958154] {CVE-2022-24448}
- Revert "btrfs: inode: refactor the parameters of insert_reserved_file_extent()" (Srikanth C S) [Orabug: 33958240]
- Revert "btrfs: fix metadata reservation for fallocate that leads to transaction aborts" (Srikanth C S) [Orabug: 33958240]
- ovl: fix WARN_ON nlink drop to zero (Miklos Szeredi) [Orabug: 33687076]
- panic: reinitialize logbuf locks before notifiers (Stephen Brennan) [Orabug: 33740420]
- printk: Drop console_sem during panic (Stephen Brennan) [Orabug: 33740420]
- printk: Avoid livelock with heavy printk during panic (Stephen Brennan) [Orabug: 33740420]
- printk: disable optimistic spin during panic (Stephen Brennan) [Orabug: 33740420]
- printk: Add panic_in_progress helper (Stephen Brennan) [Orabug: 33740420]
- sched: Put vcpu preemption idle check into a SCHED_FEAT. (chris hyser) [Orabug: 33806261]
- rds/ib: recover rds connection from stuck rx path (Nagappan Ramasamy Palaniappan) [Orabug: 33820776]
- drm/vmwgfx: Fix stale file descriptors on failed usercopy (Mathias Krause) [Orabug: 33840432] {CVE-2022-22942}
- udf: Restore i_lenAlloc when inode expansion fails (Jan Kara) [Orabug: 33870266] {CVE-2022-0617}
- udf: Fix NULL ptr deref when converting from inline format (Jan Kara) [Orabug: 33870266] {CVE-2022-0617}
- ACPI: sysfs: copy ACPI data using io memory copying (Colin Ian King) [Orabug: 33876016]
- Enable CONFIG_DM_DUST and nano for UEK6 (Gulam Mohamed) [Orabug: 33897851]
- arm64/efi: remove spurious WARN_ON for !4K kernels (Mark Rutland) [Orabug: 33900748]
- lib/iov_iter: initialize "flags" in new pipe_buffer (Max Kellermann) [Orabug: 33910799]
- ipv4: tcp: send zero IPID in SYNACK messages (Eric Dumazet) [Orabug: 33917056] {CVE-2020-36516}
- ipv4: avoid using shared IP generator for connected sockets (Eric Dumazet) [Orabug: 33917056] {CVE-2020-36516}
- bpf: fix out-of-tree module build (Alan Maguire) [Orabug: 33919430]
- ACPICA: Enable sleep button on ACPI legacy wake (Anchal Agarwal) [Orabug: 33925471]
- arm64: Use the clearbhb instruction in mitigations (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: add ID_AA64ISAR2_EL1 sys register (Joey Gouly) [Orabug: 33921736] {CVE-2022-23960}
- KVM: arm64: Allow SMCCC_ARCH_WORKAROUND_3 to be discovered and migrated (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: Mitigate spectre style branch history side channels (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- KVM: arm64: Add templates for BHB mitigation sequences (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: Add Cortex-X2 CPU part definition (Anshuman Khandual) [Orabug: 33921736] {CVE-2022-23960}
- arm64: Add Neoverse-N2, Cortex-A710 CPU part definition (Suzuki K Poulose) [Orabug: 33921736] {CVE-2022-23960}
- arm64: Add part number for Arm Cortex-A77 (Rob Herring) [Orabug: 33921736] {CVE-2022-23960}
- arm64: proton-pack: Report Spectre-BHB vulnerabilities as part of Spectre-v2 (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: Add percpu vectors for EL1 (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Add macro for reading symbol addresses from the trampoline (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Add vectors that have the bhb mitigation sequences (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Add non-kpti __bp_harden_el1_vectors for mitigations (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Allow the trampoline text to occupy multiple pages (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Make the kpti trampoline's kpti sequence optional (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Move trampoline macros out of ifdef'd section (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Don't assume tramp_vectors is the start of the vectors (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Allow tramp_alias to access symbols after the 4K boundary (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Move the trampoline data page before the text page (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Free up another register on kpti's tramp_exit path (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry: Make the trampoline cleanup optional (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- arm64: entry.S: Add ventry overflow sanity checks (James Morse) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: Add initial retpoline support" (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: asm: Use *_nospec variants for blr and br." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: Add MIDR_APM_POTENZA." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: insn: Add offset getter/setter for adr." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: alternatives: Add support for adr/adrp with offset in alt block." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "BACKPORT: VARIANT 2: arm64: Use alternative framework for retpoline." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "Arm64: add retpoline to cpu_show_spectre_v2" (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "arm64: retpoline: Don't use retpoline in KVM's HYP part." (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "uek-rpm: aarch64 config enable RETPOLINE" (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- Revert "uek-rpm: aarch64 config enable RETPOLINE OL8" (Russell King) [Orabug: 33921736] {CVE-2022-23960}
- x86/speculation: Add knob for eibrs_retpoline_enabled (Patrick Colp) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Extend our code to properly support eibrs+lfence and eibrs+retpoline (Patrick Colp) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Update link to AMD speculation whitepaper (Kim Phillips) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Use generic retpoline by default on AMD (Kim Phillips) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Include unprivileged eBPF status in Spectre v2 mitigation reporting (Josh Poimboeuf) [Orabug: 33941936] {CVE-2021-26401}
- Documentation/hw-vuln: Update spectre doc (Peter Zijlstra) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Add eIBRS + Retpoline options (Peter Zijlstra) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Rename RETPOLINE_AMD to RETPOLINE_LFENCE (Peter Zijlstra (Intel)) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: The choice of retpoline mode is sometimes ignored (Patrick Colp) [Orabug: 33917127]
- x86/speculation: Merge one test in spectre_v2_user_select_mitigation() (Borislav Petkov) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Update ALTERNATIVEs to (more closely) match upstream (Patrick Colp) [Orabug: 33941936] {CVE-2021-26401}
- x86/speculation: Fix bug in retpoline mode on AMD with `spectre_v2=none` (Patrick Colp) [Orabug: 33941936] {CVE-2021-26401}
- ovl: fix WARN_ON nlink drop to zero (Miklos Szeredi) [Orabug: 33687076]
- panic: reinitialize logbuf locks before notifiers (Stephen Brennan) [Orabug: 33740420]
- printk: Drop console_sem during panic (Stephen Brennan) [Orabug: 33740420]
- printk: Avoid livelock with heavy printk during panic (Stephen Brennan) [Orabug: 33740420]
- printk: disable optimistic spin during panic (Stephen Brennan) [Orabug: 33740420]
- printk: Add panic_in_progress helper (Stephen Brennan) [Orabug: 33740420]
- sched: Put vcpu preemption idle check into a SCHED_FEAT. (chris hyser) [Orabug: 33806261]
- rds/ib: recover rds connection from stuck rx path (Nagappan Ramasamy Palaniappan) [Orabug: 33820776]
- drm/vmwgfx: Fix stale file descriptors on failed usercopy (Mathias Krause) [Orabug: 33840432] {CVE-2022-22942}
- udf: Restore i_lenAlloc when inode expansion fails (Jan Kara) [Orabug: 33870266] {CVE-2022-0617}
- udf: Fix NULL ptr deref when converting from inline format (Jan Kara) [Orabug: 33870266] {CVE-2022-0617}
- ACPI: sysfs: copy ACPI data using io memory copying (Colin Ian King) [Orabug: 33876016]
- Enable CONFIG_DM_DUST and nano for UEK6 (Gulam Mohamed) [Orabug: 33897851]
- arm64/efi: remove spurious WARN_ON for !4K kernels (Mark Rutland) [Orabug: 33900748]
- lib/iov_iter: initialize "flags" in new pipe_buffer (Max Kellermann) [Orabug: 33910799]
- ipv4: tcp: send zero IPID in SYNACK messages (Eric Dumazet) [Orabug: 33917056] {CVE-2020-36516}
- ipv4: avoid using shared IP generator for connected sockets (Eric Dumazet) [Orabug: 33917056] {CVE-2020-36516}
- bpf: fix out-of-tree module build (Alan Maguire) [Orabug: 33919430]
- ACPICA: Enable sleep button on ACPI legacy wake (Anchal Agarwal) [Orabug: 33925471]
- arm64: Use the clearbhb instruction in mitigations (James Morse) [Orabug: 33921736]
- arm64: add ID_AA64ISAR2_EL1 sys register (Joey Gouly) [Orabug: 33921736]
- KVM: arm64: Allow SMCCC_ARCH_WORKAROUND_3 to be discovered and migrated (James Morse) [Orabug: 33921736]
- arm64: Mitigate spectre style branch history side channels (James Morse) [Orabug: 33921736]
- KVM: arm64: Add templates for BHB mitigation sequences (James Morse) [Orabug: 33921736]
- arm64: Add Cortex-X2 CPU part definition (Anshuman Khandual) [Orabug: 33921736]
- arm64: Add Neoverse-N2, Cortex-A710 CPU part definition (Suzuki K Poulose) [Orabug: 33921736]
- arm64: Add part number for Arm Cortex-A77 (Rob Herring) [Orabug: 33921736]
- arm64: proton-pack: Report Spectre-BHB vulnerabilities as part of Spectre-v2 (James Morse) [Orabug: 33921736]
- arm64: Add percpu vectors for EL1 (James Morse) [Orabug: 33921736]
- arm64: entry: Add macro for reading symbol addresses from the trampoline (James Morse) [Orabug: 33921736]
- arm64: entry: Add vectors that have the bhb mitigation sequences (James Morse) [Orabug: 33921736]
- arm64: entry: Add non-kpti __bp_harden_el1_vectors for mitigations (James Morse) [Orabug: 33921736]
- arm64: entry: Allow the trampoline text to occupy multiple pages (James Morse) [Orabug: 33921736]
- arm64: entry: Make the kpti trampoline's kpti sequence optional (James Morse) [Orabug: 33921736]
- arm64: entry: Move trampoline macros out of ifdef'd section (James Morse) [Orabug: 33921736]
- arm64: entry: Don't assume tramp_vectors is the start of the vectors (James Morse) [Orabug: 33921736]
- arm64: entry: Allow tramp_alias to access symbols after the 4K boundary (James Morse) [Orabug: 33921736]
- arm64: entry: Move the trampoline data page before the text page (James Morse) [Orabug: 33921736]
- arm64: entry: Free up another register on kpti's tramp_exit path (James Morse) [Orabug: 33921736]
- arm64: entry: Make the trampoline cleanup optional (James Morse) [Orabug: 33921736]
- arm64: entry.S: Add ventry overflow sanity checks (James Morse) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: Add initial retpoline support" (Russell King) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: asm: Use *_nospec variants for blr and br." (Russell King) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: Add MIDR_APM_POTENZA." (Russell King) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: insn: Add offset getter/setter for adr." (Russell King) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: alternatives: Add support for adr/adrp with offset in alt block." (Russell King) [Orabug: 33921736]
- Revert "BACKPORT: VARIANT 2: arm64: Use alternative framework for retpoline." (Russell King) [Orabug: 33921736]
- Revert "Arm64: add retpoline to cpu_show_spectre_v2" (Russell King) [Orabug: 33921736]
- Revert "arm64: retpoline: Don't use retpoline in KVM's HYP part." (Russell King) [Orabug: 33921736]
- Revert "uek-rpm: aarch64 config enable RETPOLINE" (Russell King) [Orabug: 33921736]
- Revert "uek-rpm: aarch64 config enable RETPOLINE OL8" (Russell King) [Orabug: 33921736]
- x86/speculation: Add knob for eibrs_retpoline_enabled (Patrick Colp) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Extend our code to properly support eibrs+lfence and eibrs+retpoline (Patrick Colp) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Update link to AMD speculation whitepaper (Kim Phillips) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Use generic retpoline by default on AMD (Kim Phillips) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Include unprivileged eBPF status in Spectre v2 mitigation reporting (Josh Poimboeuf) [Orabug: 33941936] {CVE-2021-26341}
- Documentation/hw-vuln: Update spectre doc (Peter Zijlstra) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Add eIBRS + Retpoline options (Peter Zijlstra) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Rename RETPOLINE_AMD to RETPOLINE_LFENCE (Peter Zijlstra (Intel)) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: The choice of retpoline mode is sometimes ignored (Patrick Colp) [Orabug: 33917127]
- x86/speculation: Merge one test in spectre_v2_user_select_mitigation() (Borislav Petkov) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Update ALTERNATIVEs to (more closely) match upstream (Patrick Colp) [Orabug: 33941936] {CVE-2021-26341}
- x86/speculation: Fix bug in retpoline mode on AMD with `spectre_v2=none` (Patrick Colp) [Orabug: 33941936] {CVE-2021-26341}
- rds/ib: handle posted ACK during connection shutdown (Nagappan Ramasamy Palaniappan) [Orabug: 33820760]
- rds/ib: reap tx completions during connection shutdown (Nagappan Ramasamy Palaniappan) [Orabug: 33820760]
- rds/ib: recover rds connection from stuck tx path (Nagappan Ramasamy Palaniappan) [Orabug: 33820760]
_______________________________________________
An unbreakable Enterprise kernel security update has been released for Oracle Linux 8.