Re: [openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-02-01 Thread Hal Rosenstock
Hi again Yevgeny, On Wed, 2007-01-31 at 17:19, Yevgeny Kliteynik wrote: [snip...] + for (i = 0; i IB_MAX_NUM_VLS; i++) + { + if (valid_sls[i]) + { + vl = ib_slvl_table_get(p_slvl_tbl,i); + if (vl == IB_DROP_VL) Does vl Operational VLs need

Re: [openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-01-31 Thread Hal Rosenstock
Hi Yevgeny, On Tue, 2007-01-30 at 10:33, Yevgeny Kliteynik wrote: Checking PathRecord query for QoS constraints The QoS-aware path selection logic is implemented in a separate function that is called only when QoS in OpenSM is on. It causes some code duplication, but the idea is to

Re: [openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-01-31 Thread Sasha Khapyorsky
On 17:33 Tue 30 Jan , Yevgeny Kliteynik wrote: Checking PathRecord query for QoS constraints The QoS-aware path selection logic is implemented in a separate function that is called only when QoS in OpenSM is on. It causes some code duplication, but the idea is to minimize the changes in

Re: [openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-01-31 Thread Yevgeny Kliteynik
Hi Hal, Hal Rosenstock wrote: Hi Yevgeny, On Tue, 2007-01-30 at 10:33, Yevgeny Kliteynik wrote: Checking PathRecord query for QoS constraints The QoS-aware path selection logic is implemented in a separate function that is called only when QoS in OpenSM is on. It causes some code

Re: [openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-01-31 Thread Yevgeny Kliteynik
Hi Sasha, Sasha Khapyorsky wrote: On 17:33 Tue 30 Jan , Yevgeny Kliteynik wrote: Checking PathRecord query for QoS constraints The QoS-aware path selection logic is implemented in a separate function that is called only when QoS in OpenSM is on. It causes some code duplication, but the

[openib-general] [PATCH 10/10] osm: QoS in OpenSM

2007-01-30 Thread Yevgeny Kliteynik
Checking PathRecord query for QoS constraints The QoS-aware path selection logic is implemented in a separate function that is called only when QoS in OpenSM is on. It causes some code duplication, but the idea is to minimize the changes in the existing logic in OSM. Eventually, these two