| Message ID | 20190108110804.01e3ad9ef85ea7446ad4c61a@gmx.com |
|---|---|
| State | New |
| Headers | show |
2019-01-08 10:08 GMT+01:00, Lauri Kasanen <cand@gmx.com>: > The existing code was in no released kernel that I can see. The corrected > code > was added in 3.9. > > Signed-off-by: Lauri Kasanen <cand@gmx.com> > --- > libavutil/ppc/cpu.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/libavutil/ppc/cpu.c b/libavutil/ppc/cpu.c > index 7bb7cd8..b022149 100644 > --- a/libavutil/ppc/cpu.c > +++ b/libavutil/ppc/cpu.c > @@ -93,13 +93,13 @@ int ff_get_cpu_flags_ppc(void) > if (buf[i + 1] & PPC_FEATURE_HAS_VSX) > ret |= AV_CPU_FLAG_VSX; > #endif > -#ifdef PPC_FEATURE_ARCH_2_07 > - if (buf[i + 1] & PPC_FEATURE_HAS_POWER8) > - ret |= AV_CPU_FLAG_POWER8; > -#endif > if (ret & AV_CPU_FLAG_VSX) > av_assert0(ret & AV_CPU_FLAG_ALTIVEC); > - goto out; This seems like an unrelated change. Carl Eugen
On Wed, 9 Jan 2019 21:55:30 +0100 Carl Eugen Hoyos <ceffmpeg@gmail.com> wrote: > 2019-01-08 10:08 GMT+01:00, Lauri Kasanen <cand@gmx.com>: > > The existing code was in no released kernel that I can see. The corrected > > code > > was added in 3.9. > > > > Signed-off-by: Lauri Kasanen <cand@gmx.com> > > --- > > libavutil/ppc/cpu.c | 10 +++++----- > > 1 file changed, 5 insertions(+), 5 deletions(-) > > > > diff --git a/libavutil/ppc/cpu.c b/libavutil/ppc/cpu.c > > index 7bb7cd8..b022149 100644 > > --- a/libavutil/ppc/cpu.c > > +++ b/libavutil/ppc/cpu.c > > @@ -93,13 +93,13 @@ int ff_get_cpu_flags_ppc(void) > > if (buf[i + 1] & PPC_FEATURE_HAS_VSX) > > ret |= AV_CPU_FLAG_VSX; > > #endif > > -#ifdef PPC_FEATURE_ARCH_2_07 > > - if (buf[i + 1] & PPC_FEATURE_HAS_POWER8) > > - ret |= AV_CPU_FLAG_POWER8; > > -#endif > > if (ret & AV_CPU_FLAG_VSX) > > av_assert0(ret & AV_CPU_FLAG_ALTIVEC); > > > - goto out; > > This seems like an unrelated change. It's necessary. HWCAP appears before HWCAP2 in the array, so if the code jumps out in HWCAP, it never gets to checking the CAP2 bits like power8. - Lauri
2019-01-10 10:45 GMT+01:00, Lauri Kasanen <cand@gmx.com>: > On Wed, 9 Jan 2019 21:55:30 +0100 > Carl Eugen Hoyos <ceffmpeg@gmail.com> wrote: > >> 2019-01-08 10:08 GMT+01:00, Lauri Kasanen <cand@gmx.com>: >> > The existing code was in no released kernel that I can see. The >> > corrected >> > code >> > was added in 3.9. >> > >> > Signed-off-by: Lauri Kasanen <cand@gmx.com> >> > --- >> > libavutil/ppc/cpu.c | 10 +++++----- >> > 1 file changed, 5 insertions(+), 5 deletions(-) >> > >> > diff --git a/libavutil/ppc/cpu.c b/libavutil/ppc/cpu.c >> > index 7bb7cd8..b022149 100644 >> > --- a/libavutil/ppc/cpu.c >> > +++ b/libavutil/ppc/cpu.c >> > @@ -93,13 +93,13 @@ int ff_get_cpu_flags_ppc(void) >> > if (buf[i + 1] & PPC_FEATURE_HAS_VSX) >> > ret |= AV_CPU_FLAG_VSX; >> > #endif >> > -#ifdef PPC_FEATURE_ARCH_2_07 >> > - if (buf[i + 1] & PPC_FEATURE_HAS_POWER8) >> > - ret |= AV_CPU_FLAG_POWER8; >> > -#endif >> > if (ret & AV_CPU_FLAG_VSX) >> > av_assert0(ret & AV_CPU_FLAG_ALTIVEC); >> >> > - goto out; >> >> This seems like an unrelated change. > > It's necessary. HWCAP appears before HWCAP2 in the array, so if the > code jumps out in HWCAP, it never gets to checking the CAP2 bits like > power8. The next line (that I unfortunately cut) is: } else if (buf[i] == AT_HWCAP2) { indicating afaict that it is only reached if buf[i] is not equal to HWCAP. What do I miss? Carl Eugen
On Thu, 10 Jan 2019 18:09:21 +0100 Carl Eugen Hoyos <ceffmpeg@gmail.com> wrote: > >> > - goto out; > >> > >> This seems like an unrelated change. > > > > It's necessary. HWCAP appears before HWCAP2 in the array, so if the > > code jumps out in HWCAP, it never gets to checking the CAP2 bits like > > power8. > > The next line (that I unfortunately cut) is: > } else if (buf[i] == AT_HWCAP2) { > indicating afaict that it is only reached if buf[i] is not equal > to HWCAP. > What do I miss? The surrounding context is a loop over all bytes: for (i = 0; i < count / sizeof(*buf); i += 2) { While the out: label is after the loop. - Lauri
2019-01-10 18:21 GMT+01:00, Lauri Kasanen <cand@gmx.com>: > On Thu, 10 Jan 2019 18:09:21 +0100 > Carl Eugen Hoyos <ceffmpeg@gmail.com> wrote: > >> >> > - goto out; >> >> >> >> This seems like an unrelated change. >> > >> > It's necessary. HWCAP appears before HWCAP2 in the array, so if the >> > code jumps out in HWCAP, it never gets to checking the CAP2 bits like >> > power8. >> >> The next line (that I unfortunately cut) is: >> } else if (buf[i] == AT_HWCAP2) { >> indicating afaict that it is only reached if buf[i] is not equal >> to HWCAP. >> What do I miss? > > The surrounding context is a loop over all bytes: > for (i = 0; i < count / sizeof(*buf); i += 2) { > > While the out: label is after the loop. Thank you! Carl Eugen
On Tue, 8 Jan 2019 11:08:04 +0200 Lauri Kasanen <cand@gmx.com> wrote: > The existing code was in no released kernel that I can see. The corrected code > was added in 3.9. > > Signed-off-by: Lauri Kasanen <cand@gmx.com> > --- > libavutil/ppc/cpu.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) Ping. - Lauri
On Thu, 17 Jan 2019 09:40:09 +0200 Lauri Kasanen <cand@gmx.com> wrote: > On Tue, 8 Jan 2019 11:08:04 +0200 > Lauri Kasanen <cand@gmx.com> wrote: > > > The existing code was in no released kernel that I can see. The corrected code > > was added in 3.9. > > > > Signed-off-by: Lauri Kasanen <cand@gmx.com> > > --- > > libavutil/ppc/cpu.c | 10 +++++----- > > 1 file changed, 5 insertions(+), 5 deletions(-) > > Ping. Ping. Carl Eugen, you were the only one who looked at it - could you apply it? Given the low interest in power patches, should I be applying for commit rights? - Lauri
2019-01-27 9:32 GMT+01:00, Lauri Kasanen <cand@gmx.com>: > Given the low interest in power patches, should I be applying for > commit rights? Definitely! Carl Eugen
On Tue, 8 Jan 2019 11:08:04 +0200 Lauri Kasanen <cand@gmx.com> wrote: > The existing code was in no released kernel that I can see. The corrected code > was added in 3.9. > > Signed-off-by: Lauri Kasanen <cand@gmx.com> > --- > libavutil/ppc/cpu.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) Applying. - Lauri
diff --git a/libavutil/ppc/cpu.c b/libavutil/ppc/cpu.c index 7bb7cd8..b022149 100644 --- a/libavutil/ppc/cpu.c +++ b/libavutil/ppc/cpu.c @@ -93,13 +93,13 @@ int ff_get_cpu_flags_ppc(void) if (buf[i + 1] & PPC_FEATURE_HAS_VSX) ret |= AV_CPU_FLAG_VSX; #endif -#ifdef PPC_FEATURE_ARCH_2_07 - if (buf[i + 1] & PPC_FEATURE_HAS_POWER8) - ret |= AV_CPU_FLAG_POWER8; -#endif if (ret & AV_CPU_FLAG_VSX) av_assert0(ret & AV_CPU_FLAG_ALTIVEC); - goto out; + } else if (buf[i] == AT_HWCAP2) { +#ifdef PPC_FEATURE2_ARCH_2_07 + if (buf[i + 1] & PPC_FEATURE2_ARCH_2_07) + ret |= AV_CPU_FLAG_POWER8; +#endif } } }
The existing code was in no released kernel that I can see. The corrected code was added in 3.9. Signed-off-by: Lauri Kasanen <cand@gmx.com> --- libavutil/ppc/cpu.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)