Re: Re [PATCH] pm8001: support HDA (flashless) mode

[Mark Salyzyn <mark_salyzyn@xxxxxxxxxxx>]

If you look at the pm8001_chip_in_hda_mode test, it actually is doing some early stages of 'normal' mode initialization. The adapter in HDA mode fails at the mpi_uninit_check() (one of the first 'normal' mode operations), the remaining tests are additional stages, but in practice we actually have only seen the first one mpi_uninit_check() fail and thus switch to HDA mode by reporting the assumption that the adapter (must/may?) be in hda mode. 

There is a catch-22 in the later parts of the test, in that an Adapter in HDA mode is not doing anything, so it certainly can not report it is in this mode ;-/

Sincerely -- Mark Salyzyn

On Apr 28, 2012, at 12:44 AM, Jack Wang wrote:

> -	PM8001_CHIP_DISP->chip_soft_rst(pm8001_ha, 0x252acbcd);
> +
> +	/* HDA SEEPROM Force HDA Mode */
> +	if (PM8001_CHIP_DISP->chip_in_hda_mode(pm8001_ha)) {
> +		rc = PM8001_CHIP_DISP->chip_hda_mode(pm8001_ha);
> +		if (!rc) {
> +			rc = -EBUSY;
> +			goto err_out_ha_free;
> +		}
> +		pm8001_ha->rst_signature = SPC_HDASOFT_RESET_SIGNATURE;
> +	} else {
> +		PM8001_CHIP_DISP->chip_soft_rst(pm8001_ha,
> +			SPC_SOFT_RESET_SIGNATURE);
> +		pm8001_ha->rst_signature = SPC_SOFT_RESET_SIGNATURE;
> +	}
> 
> Here I think we'd better try normal mode first, then fail to HAD mode.
> 
> Jack
> 
>> 
>> The pm8001 can be delivered as a standalone product with flash-programmed
>> firmware images, or without the flash present requiring the driver to
> upload
>> the images into the chip's RAM and then run. This is called HDA mode.
>> 
>> We add support for this firmware upload in the enclosed patch. We try some
>> basic initialization checks of the Firmware, and if it appears dead, we
> make
>> the assumption the adapter must in-fact be halted in this HDA mode. The
>> Firmware images themselves have not been cleared for open-release by PMC,
> but
>> they are available in OpenSolaris <hint hint>. PMC's rationalization for
> not
>> wanting an open-release of the Firmware Images is that they do not want to
>> take support calls except from paying OEMs (such as Xyratex) that are
> embedding
>> PMC product into the motherboards and thus may have a tested combination
> of
>> Firmware and Hardware. Please respect this sentiment. Images are expected
> in:
>> 
>> /lib/firmware/aap1img.bin
>> /lib/firmware/ilaimg.bin
>> /lib/firmware/iopimg.bin
>> /lib/firmware/istrimg.bin
>> 
>> using the exact same naming convention as PMC and in OpenSolaris (and its
>> followon children) for these image files.
>> 
>> Signed-off-by: Mark Salyzyn <mark_salyzyn@xxxxxxxxxxx>
>> 
>> drivers/scsi/pm8001/pm8001_hwi.c  |  584
>> +++++++++++++++++++++++++++++++++++---
>> drivers/scsi/pm8001/pm8001_hwi.h  |   37 ++
>> drivers/scsi/pm8001/pm8001_init.c |   30 +
>> drivers/scsi/pm8001/pm8001_sas.h  |    3
>> 4 files changed, 613 insertions(+), 41 deletions(-)
>> 
>> Please see enclosed attachment
> 
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html