QEMU 11.1 expands what software can exercise without a matching physical board. That is valuable for correctness tests, but an emulated feature does not reproduce every timing characteristic of real hardware.
What the release adds
The August 11 release announcement includes UFS Write Booster and Host-Initiated Defragmentation emulation, big-endian RISC-V support, and nested virtualization on the Arm HVF accelerator. The full changelog also records device and guest-to-host fixes, including 9pfs and USB issues. Applicability depends on the devices actually exposed by a VM.
A UFS command path in an emulator can help exercise a guest driver’s handling of supported protocol behavior. It should not be presented as a faithful benchmark of a phone’s flash cells, cache exhaustion or thermal throttling. Those physical properties require suitable hardware evidence.
Four bytes expose a portability bug
Our example encodes the integer 0x12345678. In increasing memory-address order, its four bytes are 12 34 56 78 for big-endian storage and 78 56 34 12 for little-endian storage.
A file format or network protocol must define which order it uses. Code that copies the host’s raw integer representation into a file can therefore produce different bytes on different targets. A test using a distinct-byte value like this reveals a mistake that an all-zero value would hide.
The cover shows both byte sequences on the same address axis. It is a language-independent illustration of byte order, not a claim that QEMU makes an incompatible binary portable automatically. The guest binary, operating system and emulated target still need to match.
Nested virtualization needs host support
The QEMU 11.1 HVF implementation checks Apple’s EL2 capability API and rejects a requested nested configuration when unsupported. Its capability check is gated on macOS 15 or later. Merely owning an Apple silicon Mac is insufficient evidence that the required host capability is available.
A nested test has at least three levels: the physical host, a first guest with virtualization support, and the inner guest. Record the accelerator and configuration at each boundary. Running an ordinary guest successfully tests only the first transition.
For deployment, review the distribution’s packaged fixes and your actual device exposure. An upstream feature list helps choose a test environment; it does not determine whether a particular production installation is patched or compatible.
Separate functional device emulation from physical performance; explain byte order with a checked example and qualify HVF nested virtualization by host capability.