| *** tpb has joined #timvideos | 00:00 | |
| *** paradisaeidae has joined #timvideos | 00:31 | |
| *** CarlFK has joined #timvideos | 00:54 | |
| *** ChanServ sets mode: +v CarlFK | 00:54 | |
| *** paradisaeidae has quit IRC | 01:46 | |
| *** rohitksingh has quit IRC | 02:04 | |
| *** rohitksingh_work has joined #timvideos | 03:32 | |
| *** sb0 has quit IRC | 05:27 | |
| *** sb0 has joined #timvideos | 05:30 | |
| shorne | https://lwn.net/Articles/132804/ | 08:38 |
|---|---|---|
| tpb | Title: Big-endian I/O memory [LWN.net] (at lwn.net) | 08:38 |
| shorne | This article is a bit interesting, part of my fix for uart is to use ioread32be, but this says that most hardware should be little endian. | 08:39 |
| shorne | _florent_: mithro: is this as expected with the litex components? | 08:40 |
| shenki | ah. that's why things were a bit strange | 08:41 |
| shenki | all of the kernel's accessors assume LE: readb, ioread32, etc | 08:42 |
| shenki | oh, i wonder if that's why my ethernet driver is not working | 08:55 |
| shenki | :D | 08:56 |
| _florent_ | shorne/shenki: peripherals have CSR registers that are exposed on the Wishbone bus | 08:57 |
| _florent_ | but a CSR is only 8 bits of a Wishbone access (ie 8 bits per 32 bits) | 08:57 |
| _florent_ | the 8 bits of the CSR are on the LSB | 08:58 |
| _florent_ | so for CSR that are on more that 8 bits, let's say N*8 bits you have to do N wishbone accesses | 08:59 |
| _florent_ | you have to look at the generated csr.h to know the size of the CSR | 08:59 |
| *** fumblehool has joined #timvideos | 09:14 | |
| shorne | _florent_: I see, I noticed for some accesses byte access was fine, but for some wider accesses it required using ioread32be. | 09:15 |
| shorne | shenki: I guess for us (at lease on openrisc) we can stick with ioread/write32be. | 09:15 |
| shenki | shorne: ok | 09:30 |
| shenki | _florent_: thanks for explaining | 09:31 |
| shenki | we need something to reset now that we have console | 09:31 |
| shenki | # reboot -f | 09:31 |
| shenki | reboot: Restarting system | 09:31 |
| shenki | *** MACHINE RESTART *** | 09:31 |
| shenki | # | 09:31 |
| shenki | _florent_: any ideas on how to debug the mdio reads? | 11:10 |
| shenki | _florent_: the hdmi2usb firmware gets back all Fs | 11:10 |
| *** rohitksingh_work has quit IRC | 12:38 | |
| *** fumblehool has quit IRC | 12:53 | |
| *** rohitksingh has joined #timvideos | 13:32 | |
| *** Guest71369 is now known as skay | 13:38 | |
| *** skay has quit IRC | 13:38 | |
| *** skay has joined #timvideos | 13:38 | |
| *** ChanServ sets mode: +v skay | 13:38 | |
| *** rohitksingh has quit IRC | 14:55 | |
| *** rohitksingh has joined #timvideos | 14:57 | |
| *** rohitksingh has quit IRC | 15:17 | |
| *** rohitksingh has joined #timvideos | 15:19 | |
| *** rohitksingh has quit IRC | 15:41 | |
| *** rohitksingh has joined #timvideos | 15:44 | |
| *** rohitksingh has quit IRC | 18:47 | |
| *** CarlFK has quit IRC | 21:48 | |
| shorne | shenki: yeah, reboot is currently implemented by a special l.nop in simulators, nothing was ever put in place in hardware | 21:55 |
| shorne | I discussed it a bit earlier this user, to use the Power Management modes to trigger reboot | 21:56 |
| shorne | But it needs to more discussion | 21:56 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!