-
pwp authored
Modified Files: sctdaq/makefile.inc sctdaq/BILED/biled_lib.c sctdaq/BILED/biled_local.h sctdaq/CLOAC/cloac_lib.c sctdaq/CLOAC/makefile sctdaq/MUSTARD/mus_lib.c sctdaq/MUSTARD/mus_local.h sctdaq/SCTLV/sctlv_lib.c sctdaq/SCTLV/sctlv_lib.h sctdaq/SLOG/slog_lib.c sctdaq/SLOG/slog_private.h sctdaq/scthv/scthv_lib.c sctdaq/scthv/scthv_lib.h sctdaq/stdll/makefile sctdaq/stdll/stdll.dsp sctdaq/stdll/stdll_cint.cpp sctdaq/stlib/sct_hardware.h sctdaq/temptress/temptress_lib.c sctdaq/temptress/temptress_lib.h Merge in JCH changes for use with ATLAS RCC library (SBC) sctdaq/optif/optif_lib.c sctdaq/optif/optif_local.h Merge in JCH changes AND raise default optif receiver threshold to 120 - makes "find the lady" easier to play! sctdaq/stlib/stlib.c sctdaq/stlib/stlib.opt 1) Review use of sleep functions. Under unix, st_sleep points at usleep (scaled to units of mS). I have added a windows specific version of usleep based on the windows API precision timer functions. With PIII processors or newer this should give timing accurate to within a few tens of microS (on test, I get a maximum latency of 5). So, under windows, st_sleep now pints to my version of usleep, scaled to mS. In parallel I have slashed some of the delay calls to more reasonable values, now that st_sleep really counts in mS. If we need shorter delays, we can call usleep directly. 2) Improve (?) BLUE_HANDBAG to ramp from 2.5 to set point in 0.2V steps, following experience from LBL. Even an unclocked module can now be powered without tripping - so in certain situations one might choose to run without BLUE_HANDBAG since this was a useful diagnostic! sctdaq/stdll/TST.cpp Rationalise calls to sleep functions. Calls to st_sleep and gSystem->Sleep have been replaced by calls to TST::Sleep - which for the time being calls st_sleep. This could change, but I thought it important to use one sleep function and one only. git-svn-id: svn+ssh://svn.cern.ch/reps/atlasupstrip/software/sctdaq/trunk@338 4525493e-7705-40b1-a816-d608a930855b
716227f6pwp authoredModified Files: sctdaq/makefile.inc sctdaq/BILED/biled_lib.c sctdaq/BILED/biled_local.h sctdaq/CLOAC/cloac_lib.c sctdaq/CLOAC/makefile sctdaq/MUSTARD/mus_lib.c sctdaq/MUSTARD/mus_local.h sctdaq/SCTLV/sctlv_lib.c sctdaq/SCTLV/sctlv_lib.h sctdaq/SLOG/slog_lib.c sctdaq/SLOG/slog_private.h sctdaq/scthv/scthv_lib.c sctdaq/scthv/scthv_lib.h sctdaq/stdll/makefile sctdaq/stdll/stdll.dsp sctdaq/stdll/stdll_cint.cpp sctdaq/stlib/sct_hardware.h sctdaq/temptress/temptress_lib.c sctdaq/temptress/temptress_lib.h Merge in JCH changes for use with ATLAS RCC library (SBC) sctdaq/optif/optif_lib.c sctdaq/optif/optif_local.h Merge in JCH changes AND raise default optif receiver threshold to 120 - makes "find the lady" easier to play! sctdaq/stlib/stlib.c sctdaq/stlib/stlib.opt 1) Review use of sleep functions. Under unix, st_sleep points at usleep (scaled to units of mS). I have added a windows specific version of usleep based on the windows API precision timer functions. With PIII processors or newer this should give timing accurate to within a few tens of microS (on test, I get a maximum latency of 5). So, under windows, st_sleep now pints to my version of usleep, scaled to mS. In parallel I have slashed some of the delay calls to more reasonable values, now that st_sleep really counts in mS. If we need shorter delays, we can call usleep directly. 2) Improve (?) BLUE_HANDBAG to ramp from 2.5 to set point in 0.2V steps, following experience from LBL. Even an unclocked module can now be powered without tripping - so in certain situations one might choose to run without BLUE_HANDBAG since this was a useful diagnostic! sctdaq/stdll/TST.cpp Rationalise calls to sleep functions. Calls to st_sleep and gSystem->Sleep have been replaced by calls to TST::Sleep - which for the time being calls st_sleep. This could change, but I thought it important to use one sleep function and one only. git-svn-id: svn+ssh://svn.cern.ch/reps/atlasupstrip/software/sctdaq/trunk@338 4525493e-7705-40b1-a816-d608a930855b
Loading