On 04.02.20 04:04, Linus Torvalds wrote: > More signs of your - or somebody elses - email scripts being broken: > > On Tue, Feb 4, 2020 at 1:34 AM Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: >> >> From: David Hildenbrand <david@xxxxxxxxxx> >> Subject: mm: factor out next_present_section_nr() >> >> Let's move it to the header and use the shorter variant from >> mm/page_alloc.c (the original one will also check >> "__highest_present_section_nr + 1", which is not necessary). While at it, >> make the section_nr in next_pfn() const. >> >> In next_pfn(), we now return section_nr_to_pfn(-1) instead of -1 once we >> exceed __highest_present_section_nr, which doesn't make a difference i= n >> the caller as it is big enough (>=3D all sane end_pfn). > > Here, look at that "i= n". It looks like it was a MIME line-break (so > "in" was MIME-encoded and turned into "i=\nn") followed by you or > David re-flowing the text without MIME-decoding it. > > And note the ">=3D" thing. It should be just ">=" but again there is > left-over crud from using MIME-encoded data without decoding it. > > I am noticing that this has apparently happened before too. And maybe > it's not you. Maybe it's David Hildenbrand that has sent you already > corrupted data. > > Doing a > > git log --grep="=3D" > > shows that this mistake has been done by others before. But it's not > _that_ common. I find 25 occurrences of that "=3D" thing in the logs > over the whole history of the kernel. > Easter eggs :) > The "=\n" thing is much harder to grep for quite that trivially, or > the other "random utf8 encoded as MIME and never decoded properly" > stuff. > > But the fact that I found at least _two_ of these cases in just this > series, and it had that broken coverletter too, makes me go "Hmm". > > I tried to look for other cases, but those two emails (both from David > Hildenbrand, soo...) were the only ones I found in this series of 67. > > I can fix them up, of course, but I really hate how somebody has some > workflow that generates this corruption. I'm a simple man - I use vim+git with barely any custom scripts (well, only a cc-cmd script shared my Michal once). I've been sending patches from the same machine for years without any such glitches. The only thing I do manually is copying the cover letter from Thunderbird + adapting it when resending a version. But as the patch descriptions themselves are messed up, that doesn't explain the story. I *suspect* this is the result of (one of many) temporary mail server issues we had at Red Hat when switching providers (although a weird one - but I remember there were weird ones). -- Thanks, David / dhildenb
