Hey Chris, The issue w/ ipmimonitoring is there is some bogus info being returned from the BMC on your motherboard.
> [ 0h] = sdr_version_major[ 4b] > [ 0h] = sdr_version_minor[ 4b] > [ 0h] = record_count[16b] My code, which requires version number > 0, is a bug I should fix. However, the record_count == 0 is definitely a bug. I will have to implement a workaround for this (as well as in the 0.6.0 code that I've been working on). Does ipmi-sensors work for you? The code in ipmi-sensors does not check the record_count ahead of time, so perhaps it'll work by being blissfully unaware :-) BTW, thanks for all your help getting to the bottom of all this. Since (it seems) Intel motherboards are the only ones with SSIF, a lot of this stuff hasn't been flushed out. Al On Tue, 2008-01-22 at 14:48 -0500, Chris Morgan wrote: > On Jan 22, 2008 2:47 PM, Chris Morgan <[EMAIL PROTECTED]> wrote: > > On Jan 22, 2008 2:24 PM, Al Chu <[EMAIL PROTECTED]> wrote: > > > Hey Chris, > > > > > > It looks like you can communicate w/ the BMC now, which is good. So we > > > need to figure out why the the SSIF driver is not working by default. > > > That's bug #1. I have a few ideas about it, hopefully I can deal with > > > that later. > > > > > > > [EMAIL PROTECTED]:~# ipmimonitoring --no-probing -D SSIF > > > > --driver-address=0x42 --driver-device=/dev/i2c-0 > > > > --register-spacing=1 > > > > ipmi_monitoring_sensor_readings_by_record_id: internal error > > > > > > This a more major concern. Could you run w/ --debug and see if there is > > > any extra info it outputs. > > > > > > > Hello again Al. > > > > [EMAIL PROTECTED]:~# ipmimonitoring --no-probing -D SSIF > > --driver-address=0x42 --driver-device=/dev/i2c-0 --register-spacing=1 > > --debug > > ipmi-udm.c: 627: ipmi_open_inband: errnum (9): expression failed > > (ipmi_monitoring_ipmi_communication.c, _inband_init, 176): ipmi_cmd: > > device not found > > ipmi-udm.c: 808: ipmi_close_device: errnum (15): expression failed > > > > ipmi_monitoring_sensor_readings_by_record_id: internal error > > > > Chris > > > > Please ignore this. I forgot to load i2c-dev, i2c-core and i2c-i810. > > Here is more valid output: > > [EMAIL PROTECTED]:~# ipmimonitoring --no-probing -D SSIF > --driver-address=0x42 --driver-device=/dev/i2c-0 --register-spacing=1 > --debug > ================================================================ > [ VALUE TAG NAME:LENGTH ] > ================================================================ > [ 20h] = cmd[ 8b] > ================================================================ > ================================================================ > [ VALUE TAG NAME:LENGTH ] > ================================================================ > [ 0h] = cmd[ 8b] > [ 0h] = comp_code[ 8b] > [ 0h] = sdr_version_major[ 4b] > [ 0h] = sdr_version_minor[ 4b] > [ 0h] = record_count[16b] > [ 0h] = free_space[16b] > [ 0h] = most_recent_addition_timestamp[32b] > [ 0h] = most_recent_erase_timestamp[32b] > [ 0h] = get_sdr_repository_allocation_info_command_supported[ > 1b] > [ 0h] = reserve_sdr_repository_command_supported[ 1b] > [ 0h] = partial_add_sdr_command_supported[ 1b] > [ 0h] = delete_sdr_command_supported[ 1b] > [ 0h] = reserved[ 1b] > [ 0h] = > modal_non_modal_sdr_repository_update_operation_supported[ 2b] > [ 0h] = overflow_flag[ 1b] > ================================================================ > (ipmi_monitoring_sdr_cache.c, _ipmi_monitoring_sdr_cache_retrieve, > 493): ipmi_sdr_cache_create: invalid parmaeters > ipmi_monitoring_sensor_readings_by_record_id: internal error > > Chris -- Albert Chu [EMAIL PROTECTED] 925-422-5311 Computer Scientist High Performance Systems Division Lawrence Livermore National Laboratory _______________________________________________ Freeipmi-users mailing list Freeipmi-users@gnu.org http://lists.gnu.org/mailman/listinfo/freeipmi-users
