| Message ID | 1429206940-27525-1-git-send-email-ghackmann@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Greg, On 16/04/15 17:55, Greg Hackmann wrote: > A couple of files use ffs() without explicitly including strings.h. > Some systems will pull in ffs()'s declaration through another header > anyway, but not when compiling against bionic in AOSP master. > > Signed-off-by: Greg Hackmann <ghackmann@google.com> Thanks for the patch. I'm not sure why I haven't hit this while building with Android-x86's bionic, although it's the right thing to do. I will give it a few days more for anyone to comment and I'll push it. From a quick look I cannot see libdrm in the AOSP build. Are you experimenting with a libdrm based hwcompositor again :-) Cheers Emil
On Tue, Apr 21, 2015 at 11:31 AM, Emil Velikov <emil.l.velikov@gmail.com> wrote: > > I'm not sure why I haven't hit this while building with Android-x86's > bionic, although it's the right thing to do. Maybe a difference in bionic versions? I know the bionic team made some recent (post-5.1) changes that unintentionally exposed a handful of other missing #includes in our tree. > From a quick look I cannot see libdrm in the AOSP build. Are you > experimenting with a libdrm based hwcompositor again :-) Yeah. That's where I want to end up in the long term, anyway. I've said publicly that we'd eventually like to converge with upstream on this, and I still mean it. :)
On 22/04/15 16:19, Greg Hackmann wrote: > On Tue, Apr 21, 2015 at 11:31 AM, Emil Velikov <emil.l.velikov@gmail.com> wrote: >> >> I'm not sure why I haven't hit this while building with Android-x86's >> bionic, although it's the right thing to do. > > Maybe a difference in bionic versions? I know the bionic team made > some recent (post-5.1) changes that unintentionally exposed a handful > of other missing #includes in our tree. > Could be. I've never really bothered checking the bionic (both downstream patches and version) that Android-x86 uses. >> From a quick look I cannot see libdrm in the AOSP build. Are you >> experimenting with a libdrm based hwcompositor again :-) > > Yeah. That's where I want to end up in the long term, anyway. I've > said publicly that we'd eventually like to converge with upstream on > this, and I still mean it. :) > Great glad to hear. We have a GSoC student that is working on it this summer, so things might be closer than expected. May I suggest that you ping/send patches to the list early and often. There is a lot of traffic, and most people are busy with their own hardware so patches do get missed/forgotten. Another good idea is to pop over at #dri-devel at FreeNode, and have a chat with the guys in real time - someone might be able to find a solution/steer you in the right direction :-) Cheers, Emil
2015-04-24 0:56 GMT+08:00 Emil Velikov <emil.l.velikov@gmail.com>: > On 22/04/15 16:19, Greg Hackmann wrote: >> On Tue, Apr 21, 2015 at 11:31 AM, Emil Velikov <emil.l.velikov@gmail.com> wrote: >>> >>> I'm not sure why I haven't hit this while building with Android-x86's >>> bionic, although it's the right thing to do. >> >> Maybe a difference in bionic versions? I know the bionic team made >> some recent (post-5.1) changes that unintentionally exposed a handful >> of other missing #includes in our tree. >> > Could be. I've never really bothered checking the bionic (both > downstream patches and version) that Android-x86 uses. Each Android-x86 branch is based on the respective AOSP stable release. That means bionic of kitkat-x86 is based on ASOSP kitkat-mr2.2, while lollipop-x86 is based on the lollipop-mr1 release. It may be different than the AOSP master branch.
diff --git a/nouveau/nouveau.c b/nouveau/nouveau.c index 2d95b74..22e0f80 100644 --- a/nouveau/nouveau.c +++ b/nouveau/nouveau.c @@ -30,6 +30,7 @@ #include <stdlib.h> #include <stdint.h> #include <string.h> +#include <strings.h> #include <stdbool.h> #include <assert.h> #include <errno.h> diff --git a/tests/modetest/modetest.c b/tests/modetest/modetest.c index 53dfe05..63d2059 100644 --- a/tests/modetest/modetest.c +++ b/tests/modetest/modetest.c @@ -50,6 +50,7 @@ #include <inttypes.h> #include <unistd.h> #include <string.h> +#include <strings.h> #include <errno.h> #include <sys/poll.h> #include <sys/time.h>
A couple of files use ffs() without explicitly including strings.h. Some systems will pull in ffs()'s declaration through another header anyway, but not when compiling against bionic in AOSP master. Signed-off-by: Greg Hackmann <ghackmann@google.com> --- nouveau/nouveau.c | 1 + tests/modetest/modetest.c | 1 + 2 files changed, 2 insertions(+)