> On Apr 4, 2017, at 19:21, Pedro Giffuni <p...@freebsd.org> wrote: > >> On Apr 4, 2017, at 14:48, Kyle Evans <kevan...@ksu.edu> wrote: >> >> On Tue, Apr 4, 2017 at 2:45 PM, Ngie Cooper (yaneurabeya) >> <yaneurab...@gmail.com> wrote: >> >> > On Apr 4, 2017, at 12:04, Conrad Meyer <c...@freebsd.org> wrote: >> > >> > On Tue, Apr 4, 2017 at 11:56 AM, Dimitry Andric <d...@freebsd.org> wrote: >> >> On 4 Apr 2017, at 19:14, Ngie Cooper (yaneurabeya) >> >> <yaneurab...@gmail.com> wrote: >> >>> Where did xmalloc.c originate from? >> >> >> >> GNU. >> > >> > I believe this to be completely incorrect. >> > >> >> Almost all software from the GNU project relies on malloc wrappers >> >> which abort the program on allocation failures. >> > >> > That is not what bsdgrep's xmalloc() did, if you read the code. It >> > simply tracks all allocations for basic leak analysis. >> > >> > Abort on allocation failure would be a perfectly reasonable behavior >> > for bsdgrep(1), too. >> >> There are multiple, competing definitions floating around the internet. I >> was genuinely curious where this variant came from because I wanted to make >> sure we weren’t just zapping a file that some upstream uses somewhere, in >> the event we were going to bring down further updates, again, from said >> upstream source. >> >> FWIW- I did scour the internet for other bsdgrep implementations and did not >> find a trace of this in any of the others that I found, to include the OS X >> implementation. In fact, as I recall, most of them didn't even have the >> regex/ bits that we do, presumably they were using regex(3) but it's been a >> while since I was poking around. > > Well, for the history of bsdgrep, you have to go back to freegrep: > > https://jameshoward.us/software/freegrep/ > > but the most significant changes are likely to be due to Gabor: > > https://wiki.freebsd.org/BSDgrep
So…. Gabor added xmalloc then ;p… -Ngie
signature.asc
Description: Message signed with OpenPGP using GPGMail