Bold opening: Linux 6.19 brings USB 3.0 compatibility to Apple Silicon, closing a long-standing gap for M1/M1 Pro/Max devices. Here’s what changed and why it matters.
Last night’s Linux 6.19 kernel merge window included a complete set of USB and Thunderbolt driver updates. The standout development this cycle is the arrival of working USB3 support for Apple Silicon devices—specifically M1 Macs—on the mainline Linux kernel. While USB3 functionality has existed in downstream work, notably within the Asahi Linux project, the mainline patchset now enshrines the necessary changes to enable USB3 on these modern Apple systems. The M3 and M4 Macs remain the subject of ongoing reverse engineering by the remaining Asahi developers.
The Apple Silicon USB3 patches had been under review for several months and were accepted for the upcoming kernel release. Apple devices rely on the Synopsys DesignWare Core USB 3.0 controller IP (DWC3), but implementing stable USB3 operation on this hardware required a series of specific adjustments. These adjustments, detailed in prior reporting, have now been integrated into Linux 6.19.
Contributors like Sven Peter delivered the essential components—a glue layer for the DWC3 interface on Apple Silicon, DeviceTree bindings for Apple devices, and additional related changes—which are now part of the mainline kernel. Beyond Apple, the pull also adds USB support enhancements for other platforms: the ASpeed AST2700, Qualcomm Kaanapali and Glymour Device Tree additions, MediaTek MT8189 XHCI support, and ESWIN EIC7700 USB controller support.
Why this matters: USB3 on Apple Silicon enables faster peripheral connectivity and better compatibility with a broad ecosystem of USB devices on Linux, improving the overall user experience for developers and enthusiasts running Linux on Apple hardware.
Controversy to ponder: Some users may question whether mainline USB3 support should have arrived earlier or via alternative approaches, given the ecosystem’s fragmentation between mainline, downstream patches, and distributions. Do you think mainline integration is sufficient, or should distribution developers push for even earlier inclusion and broader hardware coverage? Share your thoughts in the comments.
