On Sat, Sep 05, 2026 at 03:14:17AM +0200, Florian Westphal wrote: > Geliang Tang <[email protected]> wrote: > > [ CC Sungho Bae ] > > Ping, I still get this splat on mainline, this patch fixes things for > me, so > > Tested-by: Florian Westphal <[email protected]> > > Could someone push this to Linus please (if not yet queued up)? > > Thanks. > > > put_chars() correctly passes GFP_ATOMIC, but alloc_buf() silently > > discards it and allocates with GFP_KERNEL, which may sleep. > > > > Pass the caller-supplied gfp flag through to kmalloc_flex() so > > that atomic callers get atomic allocations. > > > > Fixes: 69050f8d6d07 ("treewide: Replace kmalloc with kmalloc_obj for > > non-scalar types") > > > Hmm, probably: > > Fixes: fc220d6be3c7 ("virtio_console: refactor __send_to_port() buffer > ownership") > > ... as that adds "gfp_t gfp", but doesn't fix up kmalloc_flex() which > has GFP_KERNEL as a hidden argument via macro expansion.
Sorry, this is on me, I've been swamped, will try to get this to him "soon". greg k-h
