Paolo Bonzini <[email protected]> writes: > On 1/14/26 11:45, Markus Armbruster wrote: >>> ---- target/i386/nvmm/nvmm-all.c.orig 2024-11-20 22:48:05.000000000 >>> +0000 >>> +--- target/i386/nvmm/nvmm-all.c.orig 2025-08-26 18:32:38.000000000 >>> +0000 >>> +++ target/i386/nvmm/nvmm-all.c >>> -@@ -1057,7 +1057,11 @@ nvmm_process_section(MemoryRegionSection >>> +@@ -984,7 +984,7 @@ nvmm_init_vcpu(CPUState *cpu) >>> + } >>> + } >>> + >>> +- qcpu->vcpu_dirty = true; >>> ++ cpu->vcpu_dirty = true; >>> + cpu->accel = qcpu; >>> + >>> + return 0; >>> +@@ -1059,7 +1059,11 @@ nvmm_process_section(MemoryRegionSection >>> unsigned int delta; >>> uintptr_t hva; >> That they didn't immediately post the fix upstream is a bit of a >> disappointment. Deep in the weeds, I guess. > > Eh, it happens. :/ For the future, since Ryo and Reinoud are in Cc, it would > have been enough to contact anyone on IRC or reply to the release > announcement.
Yes, it happens, especially when you're deep in the weeds.
