Joe Perches <joe@xxxxxxxxxxx> writes: > On Fri, 2020-07-10 at 17:14 +0200, Daniel Borkmann wrote: >> On 7/10/20 8:17 AM, Alexander A. Klimov wrote: >> > Am 09.07.20 um 22:37 schrieb Daniel Borkmann: >> > > On 7/9/20 9:42 PM, Alexander A. Klimov wrote: >> > > > Rationale: >> > > > Documentation/arm/ixp4xx.rst contains "xdp" as part of "ixdp465" >> > > > which has nothing to do with XDP. > [] >> > > > diff --git a/MAINTAINERS b/MAINTAINERS > [] >> > > > @@ -18708,8 +18708,8 @@ F: include/trace/events/xdp.h >> > > > F: kernel/bpf/cpumap.c >> > > > F: kernel/bpf/devmap.c >> > > > F: net/core/xdp.c >> > > > -N: xdp >> > > > -K: xdp >> > > > +N: (?:\b|_)xdp(?:\b|_) >> > > > +K: (?:\b|_)xdp(?:\b|_) >> > > >> > > Please also include \W to generally match on non-alphanumeric char given you >> > > explicitly want to avoid [a-z0-9] around the term xdp. >> > Aren't \W, ^ and $ already covered by \b? >> >> Ah, true; it says '\b really means (?:(?<=\w)(?!\w)|(?<!\w)(?=\w))', so all good. >> In case this goes via net or net-next tree: > > This N: pattern does not match files like: > > samples/bpf/xdp1_kern.c > > and does match files like: > > drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c > > Should it? I think the idea is that it should match both? -Toke
