| Message ID | 1465517368-12536-1-git-send-email-bill.fischofer@linaro.org |
|---|---|
| State | Accepted |
| Commit | 4cf18bb4f799e6601c20e54fafbdb58da538de89 |
| Headers | show |
Merged, Maxim. On 06/10/16 03:09, Bill Fischofer wrote: > Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the > typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with other > buffer types. This enables pool 0 to be used as a timeout pool. > > Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> > --- > platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h b/platform/linux-generic/include/odp/api/plat/timer_types.h > index 93ea162..68d6f6f 100644 > --- a/platform/linux-generic/include/odp/api/plat/timer_types.h > +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h > @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); > > typedef ODP_HANDLE_T(odp_timeout_t); > > -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) > +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0xffffffff) > > /** > * @}
Hi, ODP-DPDK uses 0 or NULL for all _INVALID constants, because that's what rte_mbuf's using for that purpose. How could this fix work there? I mean, simply just applying this patch doesn't work, because then this value will be in conflict with ODP_EVENT_INVALID et all. Zoli On 10/06/16 01:09, Bill Fischofer wrote: > Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the > typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with other > buffer types. This enables pool 0 to be used as a timeout pool. > > Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> > --- > platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h b/platform/linux-generic/include/odp/api/plat/timer_types.h > index 93ea162..68d6f6f 100644 > --- a/platform/linux-generic/include/odp/api/plat/timer_types.h > +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h > @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); > > typedef ODP_HANDLE_T(odp_timeout_t); > > -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) > +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0xffffffff) > > /** > * @} >
odp-dpdk doesn't need this patch because you're using pointers as the handle values, so NULL is always a distinguished pointer. In odp-linux, 0 is a valid buffer token, which is why this caused problems in that implementation. On Wed, Jun 15, 2016 at 10:22 PM, Zoltan Kiss <zoltan.kiss@linaro.org> wrote: > Hi, > > ODP-DPDK uses 0 or NULL for all _INVALID constants, because that's what > rte_mbuf's using for that purpose. How could this fix work there? I mean, > simply just applying this patch doesn't work, because then this value will > be in conflict with ODP_EVENT_INVALID et all. > > Zoli > > > On 10/06/16 01:09, Bill Fischofer wrote: > >> Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the >> typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with >> other >> buffer types. This enables pool 0 to be used as a timeout pool. >> >> Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> >> --- >> platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h >> b/platform/linux-generic/include/odp/api/plat/timer_types.h >> index 93ea162..68d6f6f 100644 >> --- a/platform/linux-generic/include/odp/api/plat/timer_types.h >> +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h >> @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); >> >> typedef ODP_HANDLE_T(odp_timeout_t); >> >> -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) >> +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0xffffffff) >> >> /** >> * @} >> >>
On 06/16/16 07:21, Bill Fischofer wrote: > odp-dpdk doesn't need this patch because you're using pointers as the > handle values, so NULL is always a distinguished pointer. In odp-linux, 0 > is a valid buffer token, which is why this caused problems in that > implementation. > I assume if there is pointer than: #define ODP_TIMEOUT_INVALID NULL if some platform define, than you should use it: #define ODP_TIMEOUT_INVALID SDK_INTERNAL_DEFINE_FOR_INVALID_TMO Maxim. > > On Wed, Jun 15, 2016 at 10:22 PM, Zoltan Kiss <zoltan.kiss@linaro.org> > wrote: > >> Hi, >> >> ODP-DPDK uses 0 or NULL for all _INVALID constants, because that's what >> rte_mbuf's using for that purpose. How could this fix work there? I mean, >> simply just applying this patch doesn't work, because then this value will >> be in conflict with ODP_EVENT_INVALID et all. >> >> Zoli >> >> >> On 10/06/16 01:09, Bill Fischofer wrote: >> >>> Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the >>> typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with >>> other >>> buffer types. This enables pool 0 to be used as a timeout pool. >>> >>> Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> >>> --- >>> platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- >>> 1 file changed, 1 insertion(+), 1 deletion(-) >>> >>> diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h >>> b/platform/linux-generic/include/odp/api/plat/timer_types.h >>> index 93ea162..68d6f6f 100644 >>> --- a/platform/linux-generic/include/odp/api/plat/timer_types.h >>> +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h >>> @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); >>> >>> typedef ODP_HANDLE_T(odp_timeout_t); >>> >>> -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) >>> +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0xffffffff) >>> >>> /** >>> * @} >>> >>> > _______________________________________________ > lng-odp mailing list > lng-odp@lists.linaro.org > https://lists.linaro.org/mailman/listinfo/lng-odp
On Thu, Jun 16, 2016 at 2:38 PM, Maxim Uvarov <maxim.uvarov@linaro.org> wrote: > On 06/16/16 07:21, Bill Fischofer wrote: > >> odp-dpdk doesn't need this patch because you're using pointers as the >> handle values, so NULL is always a distinguished pointer. In odp-linux, 0 >> is a valid buffer token, which is why this caused problems in that >> implementation. >> >> I assume if there is pointer than: > > #define ODP_TIMEOUT_INVALID NULL > > if some platform define, than you should use it: > > #define ODP_TIMEOUT_INVALID SDK_INTERNAL_DEFINE_FOR_INVALID_TMO No, that doesn't work because ODP uses strong types. You so if you try to say: if (timeout == ODP_BUFFER_INVALID) the compiler will flag that as a type mismatch, even though ODP_BUFFER_INVALID and ODP_TIMEOUT_INVALID use the same bit pattern. This patch simply makes ODP_TIMEOUT_INVALID use the same designated "INVALID" bit pattern as the other event types in odp-linux use. > > > > Maxim. > > >> On Wed, Jun 15, 2016 at 10:22 PM, Zoltan Kiss <zoltan.kiss@linaro.org> >> wrote: >> >> Hi, >>> >>> ODP-DPDK uses 0 or NULL for all _INVALID constants, because that's what >>> rte_mbuf's using for that purpose. How could this fix work there? I mean, >>> simply just applying this patch doesn't work, because then this value >>> will >>> be in conflict with ODP_EVENT_INVALID et all. >>> >>> Zoli >>> >>> >>> On 10/06/16 01:09, Bill Fischofer wrote: >>> >>> Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the >>>> typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with >>>> other >>>> buffer types. This enables pool 0 to be used as a timeout pool. >>>> >>>> Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> >>>> --- >>>> platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- >>>> 1 file changed, 1 insertion(+), 1 deletion(-) >>>> >>>> diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h >>>> b/platform/linux-generic/include/odp/api/plat/timer_types.h >>>> index 93ea162..68d6f6f 100644 >>>> --- a/platform/linux-generic/include/odp/api/plat/timer_types.h >>>> +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h >>>> @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); >>>> >>>> typedef ODP_HANDLE_T(odp_timeout_t); >>>> >>>> -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) >>>> +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, >>>> 0xffffffff) >>>> >>>> /** >>>> * @} >>>> >>>> >>>> _______________________________________________ >> lng-odp mailing list >> lng-odp@lists.linaro.org >> https://lists.linaro.org/mailman/listinfo/lng-odp >> > > _______________________________________________ > lng-odp mailing list > lng-odp@lists.linaro.org > https://lists.linaro.org/mailman/listinfo/lng-odp >
On 06/20/16 15:27, Bill Fischofer wrote: > > On Thu, Jun 16, 2016 at 2:38 PM, Maxim Uvarov <maxim.uvarov@linaro.org > <mailto:maxim.uvarov@linaro.org>> wrote: > > On 06/16/16 07:21, Bill Fischofer wrote: > > odp-dpdk doesn't need this patch because you're using pointers > as the > handle values, so NULL is always a distinguished pointer. In > odp-linux, 0 > is a valid buffer token, which is why this caused problems in that > implementation. > > I assume if there is pointer than: > > #define ODP_TIMEOUT_INVALID NULL > > if some platform define, than you should use it: > > #define ODP_TIMEOUT_INVALID SDK_INTERNAL_DEFINE_FOR_INVALID_TMO > > > No, that doesn't work because ODP uses strong types. You so if you try > to say: > > if (timeout == ODP_BUFFER_INVALID) > > the compiler will flag that as a type mismatch, even though > ODP_BUFFER_INVALID and ODP_TIMEOUT_INVALID use the same bit pattern. > This patch simply makes ODP_TIMEOUT_INVALID use the same designated > "INVALID" bit pattern as the other event types in odp-linux use. btw, if you are on ofp mailing list this patch also fixed some internal ofp test. Maxim. > > > > Maxim. > > > On Wed, Jun 15, 2016 at 10:22 PM, Zoltan Kiss > <zoltan.kiss@linaro.org <mailto:zoltan.kiss@linaro.org>> > wrote: > > Hi, > > ODP-DPDK uses 0 or NULL for all _INVALID constants, > because that's what > rte_mbuf's using for that purpose. How could this fix work > there? I mean, > simply just applying this patch doesn't work, because then > this value will > be in conflict with ODP_EVENT_INVALID et all. > > Zoli > > > On 10/06/16 01:09, Bill Fischofer wrote: > > Resolve bug > https://bugs.linaro.org/show_bug.cgi?id=2316 by > changing the > typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be > consistent with > other > buffer types. This enables pool 0 to be used as a > timeout pool. > > Signed-off-by: Bill Fischofer > <bill.fischofer@linaro.org > <mailto:bill.fischofer@linaro.org>> > --- > platform/linux-generic/include/odp/api/plat/timer_types.h > | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git > a/platform/linux-generic/include/odp/api/plat/timer_types.h > b/platform/linux-generic/include/odp/api/plat/timer_types.h > index 93ea162..68d6f6f 100644 > --- > a/platform/linux-generic/include/odp/api/plat/timer_types.h > +++ > b/platform/linux-generic/include/odp/api/plat/timer_types.h > @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); > > typedef ODP_HANDLE_T(odp_timeout_t); > > -#define ODP_TIMEOUT_INVALID > _odp_cast_scalar(odp_timeout_t, 0) > +#define ODP_TIMEOUT_INVALID > _odp_cast_scalar(odp_timeout_t, 0xffffffff) > > /** > * @} > > > _______________________________________________ > lng-odp mailing list > lng-odp@lists.linaro.org <mailto:lng-odp@lists.linaro.org> > https://lists.linaro.org/mailman/listinfo/lng-odp > > > _______________________________________________ > lng-odp mailing list > lng-odp@lists.linaro.org <mailto:lng-odp@lists.linaro.org> > https://lists.linaro.org/mailman/listinfo/lng-odp > >
diff --git a/platform/linux-generic/include/odp/api/plat/timer_types.h b/platform/linux-generic/include/odp/api/plat/timer_types.h index 93ea162..68d6f6f 100644 --- a/platform/linux-generic/include/odp/api/plat/timer_types.h +++ b/platform/linux-generic/include/odp/api/plat/timer_types.h @@ -36,7 +36,7 @@ typedef ODP_HANDLE_T(odp_timer_t); typedef ODP_HANDLE_T(odp_timeout_t); -#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0) +#define ODP_TIMEOUT_INVALID _odp_cast_scalar(odp_timeout_t, 0xffffffff) /** * @}
Resolve bug https://bugs.linaro.org/show_bug.cgi?id=2316 by changing the typedef of ODP_TIMEOUT_INVALID to be 0xffffffff to be consistent with other buffer types. This enables pool 0 to be used as a timeout pool. Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> --- platform/linux-generic/include/odp/api/plat/timer_types.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)