Changes in arm

Changes in arm http://nxr.netbsd.org/rss/src/sys/arch/arm/ en Copyright 2005 Java fix cvs revision comment not expanding Thu Jul 30 15:00:00 UTC 2026 yurix evbarm/am335x: initialize usb phy This fixes the USB devices not being detected part of PR port-evbarm/60404. Wed Jul 29 08:00:00 UTC 2026 yurix Does not use i2c_bitbang, so should not depend on it. Fri Jul 17 17:00:00 UTC 2026 thorpej Remove unused variable. Wed Jul 15 06:00:00 UTC 2026 martin Now that bad144 handling is isolated to the places that care about it, garbage-collect it from all of the various machine-dependent files that have been cargo-culting it around needlessly for 20+ years. Version bump to 11.99.7 because this changes the size of struct cpu_disklabel on some platforms, and that structure is exposed in the module ABI. Tue Jul 14 13:00:00 UTC 2026 thorpej Move the definion of "struct fdt_platform" to <machine/fdt_platform.h>. Sun Jul 12 05:00:00 UTC 2026 thorpej KNF comments Tue Jul 07 04:00:00 UTC 2026 skrll fix few typos in comments and error message. Sat Jul 04 17:00:00 UTC 2026 andvar Break out and expand FDT property parsing for awge(4) FDT attachment frontends This uses the device tree's possible burst length bus interface settings, and improves performance on, for instance, Rockchip RK3399 Mon Jun 22 20:00:00 UTC 2026 jakllsch fix typos in comments, mainly s/archtypal/archetypal/. Sun Jun 21 10:00:00 UTC 2026 andvar s/inconsitent/inconsistent/ in comments and panic message. Wed Jun 10 21:00:00 UTC 2026 andvar arm: relax coherent DMA ordering barriers from DSB to DMB Use DMB instead of DSB for the ARM coherent DMA ordering macros dma_*_*() The previous definitions used DSB, which enforces completion semantics and is heavier than needed for coherent device DMA ordering. DMB provides ordering of memory operations without requiring full completion, making it the appropriate barrier for these coherent-only CPU/device DMA paths. Tested on Fusion VM, Orion O6, and Thunderx. There is an approximate 1% performance improvement for the Fusion VM, but less for Orion O6 and Thunderx. Fri Jun 05 08:00:00 UTC 2026 skrll fix various typos in comments. Mon Jun 01 20:00:00 UTC 2026 andvar gicv3_its: Fix ITT sizing. The ITT being allocated did not match the size specified on the MAPD command. This could cause hardware to read past the end of the ITT. The old code used a fixed offset mapping scheme to assign eventIDs (the eventID was derived from the LPI INTID). This scheme is wasteful and doesn't scale well as the ITT is essentially an array of ITEs starting with eventID 0. This change introduces per-deviceID namespaces for eventIDs and allocates them starting with 0. A fixed number of eventIDs is made available for each deviceID (MAXCPUS * 2). On a platform with an ITE size of 8 bytes, this allows the ITT to fit in a single page. The ITT is sized larger than the initial request as multiple requestors could potentially share the same deviceID (this is the case for legacy PCI devices behind a PCIe-to-PCI bridge). The size parameter of the MAPD command now matches the allocated ITT. Tested on QEMU KVM Virtual Machine and Ampere eMAG (Lenovo HR330A). Sun May 24 22:00:00 UTC 2026 jmcneill psci: ensure cores are in a known state before SYSTEM_OFF The PSCI spec says that the caller must place all cores in a known state prior to calling SYSTEM_OFF. It doesn't really specify what a "known state" is, but it does suggest that one way to do this is to call CPU_OFF on all other cores. So let's do that. This has the nice side effect of workaround around a heartbeat timeout when attempting to power down a system with firmware that doesn't return from a SYSTEM_OFF call but does not actually know how to power off the system (like Rockchip's secure firmware on RK356x based boards). Fri May 15 22:00:00 UTC 2026 jmcneill arm/fdt: introduce a generic setstatclockrate function These changes lay the groundwork for cleaning up setstatclockrate implementations across arm timers. Mon May 11 19:00:00 UTC 2026 yurix arm/generic: use generic setstatclockrate This allows us to get rid of the weak symbol hack in a9tmr since GENERIC is the only configuration with both a9tmr and gtmr active in one configuration. Configurations without __HAVE_GENERIC_SETSTATCLOCKRATE still get a setstatclockrate from the ifdef. Mon May 11 19:00:00 UTC 2026 yurix