kernel/linux.git/arch/um/os-Linux, branch v6.19.11

um: Preserve errno within signal handler

2026-03-04T12:20:34+00:00

[ Upstream commit f68b2d5a907b53eed99cf2efcaaae116df73c298 ] We rely on errno to determine whether a syscall has failed, so we need to ensure that accessing errno is async-signal-safe. Currently, we preserve the errno in sig_handler_common(), but it doesn't cover every possible case. Let's do it in hard_handler() instead, which is the signal handler we actually register. Signed-off-by: Tiwei Bie Link: https://patch.msgid.link/20260106001228.1531146-2-tiwei.btw@antgroup.com Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

um: Always set up AT_HWCAP and AT_PLATFORM

2025-11-06T12:02:34+00:00

Historically the code to set up AT_HWCAP and AT_PLATFORM was only built for 32bit x86 as it was intermingled with the vDSO passthrough code. Now that vDSO passthrough has been removed, always pass through AT_HWCAP and AT_PLATFORM. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20251028-uml-remove-32bit-pseudo-vdso-v1-10-e930063eff5f@weissschuh.net Signed-off-by: Johannes Berg

x86/um: Do not inherit vDSO from host

2025-11-06T12:02:33+00:00

Inheriting the vDSO from the host is problematic. The values read from the time functions will not be correct for the UML kernel. Furthermore the start and end of the vDSO are not stable or detectable by userspace. Specifically the vDSO datapages start before AT_SYSINFO_EHDR and the vDSO itself is larger than a single page. This codepath is only used on 32bit x86 UML. In my testing with both 32bit and 64bit hosts the passthrough functionality has always been disabled anyways due to the checks against envp in scan_elf_aux(). Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20251028-uml-remove-32bit-pseudo-vdso-v1-4-e930063eff5f@weissschuh.net Signed-off-by: Johannes Berg

um: Split out default elf_aux_hwcap

2025-11-06T12:02:33+00:00

Setting all auxiliary vector values to default values if one of them was not provided by the host will discard perfectly fine values. Remove the elf_aux_platform fallback from the vDSO ones. As zero is the correct fallback anyways, don't create a new conditional. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20251028-uml-remove-32bit-pseudo-vdso-v1-3-e930063eff5f@weissschuh.net Signed-off-by: Johannes Berg

x86/um: Move ELF_PLATFORM fallback to x86-specific code

2025-11-06T12:02:33+00:00

The generic UM code should not have references to x86-specific value. Move the fallback into the x86-specific header. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20251028-uml-remove-32bit-pseudo-vdso-v1-2-e930063eff5f@weissschuh.net Signed-off-by: Johannes Berg

um: Split out default elf_aux_platform

2025-11-06T12:02:32+00:00

Setting all auxiliary vector values to default values if one of them was not provided by the host will discard perfectly fine values. Move the elf_aux_platform fallback to its own conditional. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20251028-uml-remove-32bit-pseudo-vdso-v1-1-e930063eff5f@weissschuh.net Signed-off-by: Johannes Berg

um: Add initial SMP support

2025-10-27T15:41:15+00:00

Add initial symmetric multi-processing (SMP) support to UML. With this support enabled, users can tell UML to start multiple virtual processors, each represented as a separate host thread. In UML, kthreads and normal threads (when running in kernel mode) can be scheduled and executed simultaneously on different virtual processors. However, the userspace code of normal threads still runs within their respective single-threaded stubs. That is, SMP support is currently available both within the kernel and across different processes, but still remains limited within threads of the same process in userspace. Signed-off-by: Tiwei Bie Link: https://patch.msgid.link/20251027001815.1666872-6-tiwei.bie@linux.dev Signed-off-by: Johannes Berg

um: Define timers on a per-CPU basis

2025-10-27T15:41:15+00:00

Define timers on a per-CPU basis to enable each CPU to have its own timer. This is a preparation for adding SMP support. Signed-off-by: Tiwei Bie Link: https://patch.msgid.link/20251027001815.1666872-5-tiwei.bie@linux.dev Signed-off-by: Johannes Berg

um: Determine sleep based on need_resched()

2025-10-27T15:41:15+00:00

With SMP and NO_HZ enabled, the CPU may still need to sleep even if the timer is disarmed. Switch to deciding whether to sleep based on pending resched. Additionally, because disabling IRQs does not block SIGALRM, it is also necessary to check for any pending timer alarms. This is a preparation for adding SMP support. Signed-off-by: Tiwei Bie Link: https://patch.msgid.link/20251027001815.1666872-4-tiwei.bie@linux.dev Signed-off-by: Johannes Berg

um: Turn signals_* into thread-local variables

2025-10-27T15:41:15+00:00

Turn signals_enabled, signals_pending and signals_active into thread-local variables. This enables us to control and track signals independently on each CPU thread. This is a preparation for adding SMP support. Signed-off-by: Tiwei Bie Link: https://patch.msgid.link/20251027001815.1666872-3-tiwei.bie@linux.dev Signed-off-by: Johannes Berg