Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1572748
| From | Ard Biesheuvel <ard.biesheuvel@linaro.org> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [PATCH] log2: make order_base_2() behave correctly on const input value zero |
| Date | 2017-02-02 22:00 +0100 |
| Message-ID | <t6wE2-ua-9@gated-at.bofh.it> (permalink) |
| References | <t6t3r-6OU-17@gated-at.bofh.it> <t6wum-qe-13@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
> On 2 Feb 2017, at 20:49, Linus Torvalds <torvalds@linux-foundation.org> wrote: > > On Thu, Feb 2, 2017 at 9:04 AM, Ard Biesheuvel > <ard.biesheuvel@linaro.org> wrote: >> >> So update order_base_2() to adhere to its own documented interface. > > Ok, looks like you screwed up the types according to the build server. > > Making the return type "unsigned long" is slightly excessive. If you > need an unsigned long to describe the log2 of a number in the kernel, > you're playing with numbers that are *way* too big. > > The old code used "ilog2()", which returns an int. > Yeah I cloned roundup_pow_of_2, and failed to spot that ilog2 returns an int. I sent out the v2 already > Which is plenty big enough of a type. > > Linus
Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread
[PATCH] log2: make order_base_2() behave correctly on const input value zero Ard Biesheuvel <ard.biesheuvel@linaro.org> - 2017-02-02 18:10 +0100
Re: [PATCH] log2: make order_base_2() behave correctly on const input value zero kbuild test robot <lkp@intel.com> - 2017-02-02 19:00 +0100
Re: [PATCH] log2: make order_base_2() behave correctly on const input value zero Linus Torvalds <torvalds@linux-foundation.org> - 2017-02-02 21:50 +0100
Re: [PATCH] log2: make order_base_2() behave correctly on const input value zero Ard Biesheuvel <ard.biesheuvel@linaro.org> - 2017-02-02 22:00 +0100
csiph-web