Tags give the ability to mark specific points in history as being important
-
devprop-5.9-rc3
c15e1bdd · ·Device properties framework fix for 5.9-rc3 Prevent the promotion of the secondary firmware node of a device to the primary one from leaking a pointer (Heikki Krogerus).
-
acpi-5.9-rc3
4f31d53c · ·ACPI fixes for 5.9-rc3 - Avoid redundant rounding to the page size in acpi_os_map_iomem() to address a recently introduced issue with the EFI memory map permission check on ARM64 (Ard Biesheuvel). - Fix acpi_release_memory() to wait until the memory mappings released by it have been really unmapped (Rafael Wysocki). - Make the ACPI driver for AMD SoCs (APD) check the return value of acpi_dev_get_property() to avoid failures in the cases when the device property under inspection is missing (Furquan Shaikh).
-
pm-5.9-rc3
ef7d9604 · ·Power management fixes for 5.9-rc3 - Make the recently added Tegra194 cpufreq driver use read_cpuid_mpir() instead of cpu_logical_map() to avoid exporting logical_cpu_map (Sumit Gupta). - Drop the automatic system wakeup event reporting for devices with pending runtime-resume requests during system-wide suspend to avoid spurious aborts of the suspend flow (Rafael Wysocki). - Fix build warning in the intel_pstate driver documentation and improve the wording in there (Randy Dunlap). - Clean up two pieces of code in the cpufreq core (Viresh Kumar).
-
xfs-5.9-fixes-1
125eac24 · ·Fixes for 5.9: - Avoid a log recovery failure for an insert range operation by rolling deferred ops incrementally instead of at the end. - Fix an off-by-one error when calculating log space reservations for anything involving an inode allocation or free. - Fix a broken shortform xattr verifier. - Ensure that the shortform xattr header padding is always initialized to zero.
-
-
-
-
-
-
-
-
-
-
-
-
4.9.230-35
65e5cc6e · · -
-
-
timekeeping-for-printk-2020-08-23
e2d977c9 · ·NMI safe accessor to get clock monotonic, boottime, realtime as a combo timestamp for printk.
-
x86-urgent-2020-08-23
6a3ea3e6 · ·A single fix for x86 which removes the RDPID usage from the paranoid entry path and unconditionally uses LSL to retrieve the CPU number. RDPID depends on MSR_TSX_AUX. KVM has an optmization to avoid expensive MRS read/writes on VMENTER/EXIT. It caches the MSR values and restores them either when leaving the run loop, on preemption or when going out to user space. MSR_TSX_AUX is part of that lazy MSR set, so after writing the guest value and before the lazy restore any exception using the paranoid entry will read the guest value and use it as CPU number to retrieve the GSBASE value for the current CPU when FSGSBASE is enabled. As RDPID is only used in that particular entry path, there is no reason to burden VMENTER/EXIT with two extra MSR writes. Remove the RDPID optimization, which is not even backed by numbers from the paranoid entry path instead.