Converting the .p files seems like a bigger effort than I'd like to commit right now. I was able to convert the .bin file to a .elf file using objcopy and ln: Maybe this will come in handy for someone: https://github.com/PocketNC/bin2elf
On Thursday, April 9, 2020 at 11:19:43 AM UTC-6, Jason Kridner wrote: > > > > On Apr 9, 2020, at 12:54 PM, John Allwine <[email protected] > <javascript:>> wrote: > > > The PRU code for hal_pru_generic is here: > https://github.com/machinekit/machinekit-hal/tree/master/src/hal/drivers/hal_pru_generic > > You'll see a number of .p files and they are assembled using pasm, the > source code for which is also in the machinekit repository: > https://github.com/machinekit/machinekit-hal/tree/master/src/hal/support/pasm > > The copyrights at the top of the pasm source are by TI, so I imagine this > was copied from somewhere else. I don't see a way to compile the pru code > down to elf format using pasm, which is what it appears I need in order to > use remoteproc. Is there another tool I can use to compile/assemble the .p > files? Alternatively, is there another tool I can use to convert the > compiled .bin file to an elf file? > > > You’ll likely need to convert the syntax to get elf. You can use pru-gcc > or clpru (TI). > > I’d recommend giving pru-gcc a shot. It is available to run on BeagleBone > AI in the Debian package feed. > > I’d further recommend just switching to C with inline assembly. > > > > On Wednesday, April 8, 2020 at 1:23:48 PM UTC-6, John Allwine wrote: >> >> The hal_pru_generic PRU code is all written in assembly and is compiled >> down to a .bin file. Copying that bin file to the firmware doesn't work. On >> the Beaglebone Black the contents of the bin file are copied to the >> appropriate address in memory in order to load it onto the PRU. What do I >> have to do to this bin file in order to be able to copy it to >> /lib/firmare/am57xx-pru1_1-fw (or one of the other 3 pru options) and then >> start it with remoteproc? >> >> On Wednesday, April 8, 2020 at 12:45:33 AM UTC-6, Jason Kridner wrote: >>> >>> On Wed, Apr 8, 2020 at 12:28 AM John Allwine <[email protected]> >>> wrote: >>> > >>> > Thanks Jason! I’ve found enough info on the pinmux to get me by for >>> now. hal_pru_generic and the PRUs seem to be the big blocker at this point. >>> I still need to get my head around the details of hal_pru_generic to know >>> if using remoteproc would work. How do I use remoteproc to run code on the >>> PRU? >>> >>> Check out >>> https://github.com/beagleboard/cloud9-examples/tree/v2020.01/BeagleBone/AI/pru >>> >>> >>> PRU start script in the Makefile is at >>> >>> https://github.com/beagleboard/cloud9-examples/blob/v2020.01/common/Makefile#L164-L167 >>> >>> >>> > >>> > > On Apr 7, 2020, at 9:57 PM, Jason Kridner <[email protected]> >>> wrote: >>> > > >>> > > On Tue, Apr 7, 2020 at 11:27 PM John Allwine < >>> [email protected]> wrote: >>> > >> >>> > >> Is there a UIO PRU driver for the Beaglebone AI? How do I enable >>> it? I'm trying to be able to talk to the PRUs using hal_pru_generic on the >>> the Beaglebone AI, which does so on the BBB using the uio_pruss module. On >>> the AI I see a uio_pruss_shmem module loaded when looking at /proc/modules, >>> is that the same thing? >>> > > >>> > > This custom little uio_pruss_shmem driver will go away as a >>> different >>> > > existing uio driver can be configured to provide uio shared memory >>> > > access without interfering with the remoteproc driver's ability to >>> > > load the PRU code. >>> > > >>> > > It is largely the same as the uio_pruss module, but the memory >>> region >>> > > shared does not include the control registers needed to load code >>> and >>> > > start/stop the processor. That can be done with the remoteproc >>> > > mechanisms. >>> > > >>> > > Will this work for you? >>> > > >>> > > BTW, sorry I never got back to you on the pinmux stuff. It keeps >>> > > dropping from my plate as I seem to always have something more >>> urgent >>> > > to do at any given moment. Can you throw a meeting time on my >>> calendar >>> > > if you still need this and we can make it a live working session? >>> The >>> > > pinmux thing isn't complicated, but I have to look around in a bunch >>> > > of places to gather the right bits. >>> > > >>> > >> >>> > >> I can run the following commands on the AI: >>> > >> >>> > >> cat /sys/class/uio/uio0/maps/map0/addr >>> > >> 0x4b200000 >>> > >> >>> > >> cat /sys/class/uio/uio0/maps/map0/size >>> > >> 0x00020000 >>> > >> >>> > >> cat /sys/class/uio/uio1/maps/map0/addr >>> > >> 0x4b280000 >>> > >> >>> > >> cat /sys/class/uio/uio1/maps/map0/size >>> > >> 0x00020000 >>> > >> >>> > >> The size on both of those seem small as the PRUSS_INTC region >>> starts at 0x00020000 according to the AM572x Technical Reference Manual on >>> page 418 (and on the Beaglebone Black the PRU uio size is 0x00080000). >>> > >> >>> > >> Anyway, I'm not very familiar with things at this level, so I may >>> not be phrasing my questions well, but any help is appreciated! >>> > >> >>> > >> -- >>> > >> For more options, visit http://beagleboard.org/discuss >>> > >> --- >>> > >> You received this message because you are subscribed to the Google >>> Groups "BeagleBoard" group. >>> > >> To unsubscribe from this group and stop receiving emails from it, >>> send an email to [email protected]. >>> > >> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/beagleboard/02eb6681-dcac-4ad2-a45c-fa27a2e346f3%40googlegroups.com. >>> >>> >>> > > >>> > > >>> > > >>> > > -- >>> > > https://beagleboard.org/about - a 501c3 non-profit educating around >>> > > open hardware computing >>> > > >>> > > -- >>> > > For more options, visit http://beagleboard.org/discuss >>> > > --- >>> > > You received this message because you are subscribed to the Google >>> Groups "BeagleBoard" group. >>> > > To unsubscribe from this group and stop receiving emails from it, >>> send an email to [email protected]. >>> > > To view this discussion on the web visit >>> https://groups.google.com/d/msgid/beagleboard/CA%2BT6QPnMrcQTc8jFkoGncpMZ18O7eSQf_Wm5Bw2KH4PMUFhsrA%40mail.gmail.com. >>> >>> >>> > >>> > -- >>> > For more options, visit http://beagleboard.org/discuss >>> > --- >>> > You received this message because you are subscribed to the Google >>> Groups "BeagleBoard" group. >>> > To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> > To view this discussion on the web visit >>> https://groups.google.com/d/msgid/beagleboard/396383D5-0B85-40C4-84D3-1AC9FAE2BF29%40allwinedesigns.com. >>> >>> >>> >> -- > For more options, visit http://beagleboard.org/discuss > --- > You received this message because you are subscribed to the Google Groups > "BeagleBoard" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected] <javascript:>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/beagleboard/6bb3d1f9-2d4c-441e-b10f-abe94017ed66%40googlegroups.com > > <https://groups.google.com/d/msgid/beagleboard/6bb3d1f9-2d4c-441e-b10f-abe94017ed66%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > -- For more options, visit http://beagleboard.org/discuss --- You received this message because you are subscribed to the Google Groups "BeagleBoard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/3262ce54-338a-443a-9bdb-4112755fb7cf%40googlegroups.com.
