remote talking to AR9271
bruce m beach
brucembeach at gmail.com
Fri Feb 26 19:33:15 PST 2016
Hello Oleksij
Heres another one.
./magpie_fw_dev/target/init/magpie.c line 126
#if MAGPIE_ENABLE_WLAN == 0
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
hif_handle = HIF_init(0);
Presumably MAGPIE_ENABLE_WLAN has never been set to 0, so
the code never got compiled, hence no compile time error.
Do you want me to bump my 1 item pending to 2 items pending
or do you want to just delete the 'aaaaaaaa' out of hand?
Bruce
On 2/25/16, bruce m beach <brucembeach at gmail.com> wrote:
>> Yea, looks buggy :) Can you provide a patch for this? Make
>> github pull request.
>
>
> Yes. I want to look at it for a few days but it looks
> completely harmless. The code in question:
>
> init.c:
> A_PRINTF("Cold reboot initiated."); <-- notice this
> #if defined(PROJECT_MAGPIE)
> iowrite32(WATCH_DOG_MAGIC_PATTERN_ADDR, 0);
> #elif defined(PROJECT_K2)
> iowrite32(MAGPIE_REG_RST_STATUS_ADDR, 0);
> #endif /* #if defined(PROJECT_MAGPIE) */
> A_USB_JUMP_BOOT();
>
> never gets executed, doesn't belong in the routine
> anyway(exception_reset()), and likely is correctly executed in
>
> usb_api_main_patch.c:
>
> void cold_reboot(void)
> {
> A_PRINTF("Cold reboot initiated."); <-- notice this
> iowrite32(MAGPIE_REG_RST_STATUS_ADDR, 0);
> A_USB_JUMP_BOOT();
> }
>
> so in the end the offending code will just get deleted. As I
> say I'll look at it for a few days, fix it and then do the
> pull request. And just for information
> MAGPIE_REG_RST_STATUS_ADDR and the K2 reset status register are
> identical address's.
>
> bruce
>
>
> On 2/24/16, Oleksij Rempel <linux at rempel-privat.de> wrote:
>> Hi Bruce,
>>
>> Am 25.02.2016 um 05:19 schrieb bruce m beach:
>>> Hello Oleksij
>>>
>>>> Complete SOC documentations is missing. We have only wifi
>>>> part of it. I created some doc here:
>>>> https://github.com/qca/open-ath9k-htc-firmware/wiki/usb-regs
>>>> https://github.com/qca/open-ath9k-htc-firmware/wiki/usb-related-issues
>>>> https://github.com/qca/open-ath9k-htc-firmware/wiki/HTC
>>>
>>>
>>> Thanks for you reply. All these things I looked at, but at
>>> the moment and for the last several months I've really been
>>> working on the build tree which hopefully I'll be finished
>>> in a month or so and can finally do some codeing. (I do have
>>> a usb user space module at this point that will send data to
>>> the 9271 but it is in an unfinished state). Right now I
>>> noticed something about the preprocessor statements in init.c
>>> which puzzles me. The only way I could see it was to copy
>>> init.c to another file and delete everything except the
>>> preprocessor statements, since the final #endif is 328 lines
>>> below the #if defined(PROJECT_MAGPIE) of interest. Below is
>>> the code section of interest:
>>>
>>> #if defined(PROJECT_MAGPIE) // outer #if
>>> ....
>>> #if defined(PROJECT_MAGPIE) // inner
>>> #elif defined(PROJECT_K2)
>>> #endif /* #if defined(PROJECT_MAGPIE) */ //inner
>>>
>>> #if SYSTEM_MODULE_PRINT
>>> #else
>>> #endif
>>>
>>> #if defined(_ROM_)
>>> #endif
>>> ....
>>> ....
>>> #endif // end of outer #if
>>>
>>> Notice that if PROJECT_MAGPIE is defined the
>>> #elif defined(PROJECT_K2)
>>> will not get executed, but if PROJECT_MAGPIE is not defined the
>>> #elif defined(PROJECT_K2) will still not get executed since
>>> it is inside the outer #if defined(PROJECT_MAGPIE) block. If I cut
>>> the whole block out a
>>>
>>> diff -s htc_9271.fw htc_9271.fw.git
>>>
>>> from my tree to the git tree are still identical.
>>
>> Yea, looks buggy :)
>> Can you provide a patch for this? Make github pull request.
>>
>>>
>>> Bruce
>>>
>>> btw I can't make out who is
>>> ath9k_htc_fw-request at lists.infradead.org
>>> should I not be sending to
>>> ath9k_htc_fw at lists.infradead.org ?
>>
>>
>> ath9k_htc_fw at lists.infradead.org is the proper one.
>>
>> --
>> Regards,
>> Oleksij
>>
>>
>
More information about the ath9k_htc_fw
mailing list