On Tue, Jun 29, 2010 at 1:49 AM, Miklos Christine <m.w.christ...@gmail.com> wrote: > Hello, > > Thank you for your thorough response. I want to disable the resets of these > registers because I sample these register values to create an estimate of > when the channel is sensed as being busy for a long periods of time. > It would make my data easier to work with. If I were to disable > ath5k_hw_ani_get_listen_time(), would it affect the performance of the > driver? If so, improving ath5k_hw_ani_get_listen_time() would be more > difficult without resetting the values since the registers overflow and we > would not know the amount of cycles the TX and RX values would have > increased until the overflow point. I think the easiest thing for me to do > would be to disable the resets, as long as it does not affect the > performance of the driver. > > Thanks, > Miklos Christine > > On Sun, Jun 27, 2010 at 6:16 PM, Bruno Randolf <b...@einfach.org> wrote: >> >> On Sat June 26 2010 05:41:10 Miklos Christine wrote: >> > Hello, >> > >> > I recently downloaded the compat-wireless-2010-06-23 version of the >> > driver. >> > I previously disabled ANI to stop the resetting of the >> > AR5K_PROFCNT_CYCLE, >> > AR5K_PROFCNT_TX, and AR5K_PROFCNT_CLR registers, but now the resets are >> > still occuring. >> > Any suggestions as to what is causing this? I want to disable just the >> > resets to those register values. >> >> hi! >> >> yes, this is caused by a recent change of mine. i did this so we can see >> busy >> times also in manual ANI mode >> (http://git.kernel.org/?p=linux/kernel/git/linville/wireless- >> testing.git;a=commit;h=9537a1623359fd24ec95ba1fe60528c70e84b2a2). feel >> free to >> revert this commit locally... >> >> may i ask you why you want to disable the resets of these registers? the >> last >> values are always stored in as->pfc_cycles, as->pfc_busy, as->pfc_tx and >> as- >> >pfc_rx, which you might be able to use in your code(?). they are updated >> roughly every second (or more often, depending on the situation). if this >> is >> not what you need, it would be possible to improve >> ath5k_hw_ani_get_listen_time() to not reset the registers. but it surely >> depends on what you want to do... >> >> bruno > > > _______________________________________________ > ath5k-devel mailing list > ath5k-devel@lists.ath5k.org > https://lists.ath5k.org/mailman/listinfo/ath5k-devel > > Hi, Miklos, I think if you want to record absolute time duration (instead of busy/free fraction), you would have to reset the registers yourself at some point eventually. The registers overflow and performs bit-shift after a while right?
YY _______________________________________________ ath5k-devel mailing list ath5k-devel@lists.ath5k.org https://lists.ath5k.org/mailman/listinfo/ath5k-devel