PRU RemoteProc documentation

GoogleGroups
1

Hi all,

I’ve done a few searches and couldn’t find any threads or “conversations” in this forum/group on the remoteproc kernel module for loading firmware onto the PRUs.

The remoteproc framework is supposed to be a Linux standard, there are some generic documents for it, but I’ve only found a general diagram and explanation in TIs SDK docs for their implementation.

Simple code to flash LEDs loads and runs fine, but larger, more complicated code that also just flashes LEDs fails to load, occasionally parts of it run, or none of it runs and the there’s a bit of a stack trace in dmesg.

The first line of the error in dmesg is:
44000000.ocp:L3 Standard Error: MASTER PRUSS2 PRU1 TARGET PCIE1 (Read): At Address: 0x00806664 : Data Access in Supervisor mode during Functional access

Is this resource_table related? Or its meaning that there’s a strange memory access?

I’m trying to find out:

  1. How to get verbose info that shows what the pru_rproc kernel module is doing
  2. How to configure it, maybe commandline, or through the sysfs
  3. Documentation of how it functions, and what it expects, such as the resource_table and firmware layout to match, how it handles errors etc.
    Thanks,

Fisher

Here’s my dmesg:

[ 76.711182] IPv6: ADDRCONF(NETDEV_CHANGE): usb0: link becomes ready
[ 76.736257] IPv6: ADDRCONF(NETDEV_CHANGE): usb1: link becomes ready
[ 77.894970] NOHZ: local_softirq_pending 08
[ 77.903344] NOHZ: local_softirq_pending 08
[ 77.923339] NOHZ: local_softirq_pending 08
[ 77.967346] NOHZ: local_softirq_pending 08
[ 77.977416] NOHZ: local_softirq_pending 08
[ 78.009731] NOHZ: local_softirq_pending 08
[ 78.227509] NOHZ: local_softirq_pending 08
[ 78.414768] NOHZ: local_softirq_pending 08
[ 78.478154] NOHZ: local_softirq_pending 08
[ 78.483369] NOHZ: local_softirq_pending 08
[ 139.787136] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 139.792152] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 139.792176] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 152.475360] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 207.719337] pru-rproc 4b2b4000.pru: can’t provide a NULL firmware
[ 234.455246] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 234.455834] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 369696
[ 234.455878] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 241.915304] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 254.522378] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 254.522557] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 254.522577] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 257.559216] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 375.451166] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 375.451886] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 369696
[ 375.451931] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 380.203249] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 387.895263] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 387.895392] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 387.895411] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 390.959144] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 407.879207] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 407.879755] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 369696
[ 407.879800] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 425.139157] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 628.915190] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 628.915736] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 369876
[ 628.915780] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 840.639171] pru-rproc 4b2b4000.pru: can’t change firmware while running
[ 845.259199] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 852.903126] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 852.903663] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 370028
[ 852.903716] ------------[ cut here ]------------
[ 852.903735] WARNING: CPU: 0 PID: 1810 at drivers/bus/omap_l3_noc.c:147 l3_interrupt_handler+0x368/0x3a4
[ 852.903744] 44000000.ocp:L3 Standard Error: MASTER PRUSS2 PRU1 TARGET PCIE1 (Read): At Address: 0x00806664 : Data Access in Supervisor mode during Functional access
[ 852.903750] Modules linked in: rpmsg_rpc rpmsg_proto pru_rproc pruss_intc bnep pruss pvrsrvkm(O) pruss_soc_bus ti_vpe ti_sc videobuf2_dma_contig ti_csc ti_vpdma v4l2_mem2mem v4l2_common videobuf2_memops videobuf2_v4l2 videobuf2_core videodev media btsdio bluetooth joydev ecdh_generic brcmfmac stmpe_adc brcmutil cfg80211 omap_remoteproc virtio_rpmsg_bus rpmsg_core usb_f_ncm usb_f_mass_storage usb_f_rndis u_ether libcomposite uio_pdrv_genirq uio cmemk(O)
[ 852.903951] CPU: 0 PID: 1810 Comm: bash Tainted: G W O 4.14.108-ti-r137 #1buster
[ 852.903958] Hardware name: Generic DRA74X (Flattened Device Tree)
[ 852.903979] [] (unwind_backtrace) from [] (show_stack+0x20/0x24)
[ 852.903994] [] (show_stack) from [] (dump_stack+0x80/0x94)
[ 852.904007] [] (dump_stack) from [] (__warn+0xec/0x114)
[ 852.904019] [] (__warn) from [] (warn_slowpath_fmt+0x58/0x74)
[ 852.904032] [] (warn_slowpath_fmt) from [] (l3_interrupt_handler+0x368/0x3a4)
[ 852.904048] [] (l3_interrupt_handler) from [] (__handle_irq_event_percpu+0x84/0x2d0)
[ 852.904061] [] (__handle_irq_event_percpu) from [] (handle_irq_event_percpu+0x3c/0x90)
[ 852.904072] [] (handle_irq_event_percpu) from [] (handle_irq_event+0x48/0x6c)
[ 852.904085] [] (handle_irq_event) from [] (handle_fasteoi_irq+0xd0/0x178)
[ 852.904098] [] (handle_fasteoi_irq) from [] (generic_handle_irq+0x34/0x44)
[ 852.904109] [] (generic_handle_irq) from [] (__handle_domain_irq+0x8c/0xfc)
[ 852.904121] [] (__handle_domain_irq) from [] (gic_handle_irq+0x4c/0x88)
[ 852.904133] [] (gic_handle_irq) from [] (__irq_svc+0x6c/0xa8)
[ 852.904140] Exception stack(0xce9bbc20 to 0xce9bbc68)
[ 852.904151] bc20: c104179c ce9bbcac 00000053 c104179f ce9bbcac c104179c ce9bbd0c d3761420
[ 852.904162] bc40: c108c08c ce9bbdac ce9bbcac ce9bbc8c ce9bbc90 ce9bbc70 c0cc6b40 c0cc2d90
[ 852.904169] bc60: 20010013 ffffffff
[ 852.904182] [] (__irq_svc) from [] (format_decode+0x5c/0x5e8)
[ 852.904194] [] (format_decode) from [] (vsnprintf+0x84/0x3f4)
[ 852.904206] [] (vsnprintf) from [] (snprintf+0x38/0x54)
[ 852.904221] [] (snprintf) from [] (dev_vprintk_emit+0x64/0x210)
[ 852.904235] [] (dev_vprintk_emit) from [] (dev_printk_emit+0x40/0x5c)
[ 852.904248] [] (dev_printk_emit) from [] (__dev_printk+0x58/0x98)
[ 852.904260] [] (__dev_printk) from [] (_dev_info+0x58/0x74)
[ 852.904276] [] (_dev_info) from [] (rproc_boot+0x574/0x634)
[ 852.904291] [] (rproc_boot) from [] (state_store+0x80/0x144)
[ 852.904305] [] (state_store) from [] (dev_attr_store+0x28/0x34)
[ 852.904319] [] (dev_attr_store) from [] (sysfs_kf_write+0x48/0x54)
[ 852.904332] [] (sysfs_kf_write) from [] (kernfs_fop_write+0xf8/0x1dc)
[ 852.904346] [] (kernfs_fop_write) from [] (__vfs_write+0x28/0x48)
[ 852.904360] [] (__vfs_write) from [] (vfs_write+0xb4/0x1c8)
[ 852.904372] [] (vfs_write) from [] (SyS_write+0x5c/0xbc)
[ 852.904386] [] (SyS_write) from [] (ret_fast_syscall+0x0/0x54)
[ 852.904393] —[ end trace 0de72da1c33b76f4 ]—
[ 852.904407] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 866.703170] pru-rproc 4b2b4000.pru: can’t change firmware while running
[ 869.891225] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 876.519176] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 876.519299] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 876.519318] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 879.759196] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 889.235104] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 889.235793] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 370028
[ 889.235848] ------------[ cut here ]------------
[ 889.235867] WARNING: CPU: 0 PID: 1810 at drivers/bus/omap_l3_noc.c:147 l3_interrupt_handler+0x368/0x3a4
[ 889.235875] 44000000.ocp:L3 Standard Error: MASTER PRUSS2 PRU1 TARGET PCIE1 (Read): At Address: 0x00806664 : Data Access in Supervisor mode during Functional access
[ 889.235881] Modules linked in: rpmsg_rpc rpmsg_proto pru_rproc pruss_intc bnep pruss pvrsrvkm(O) pruss_soc_bus ti_vpe ti_sc videobuf2_dma_contig ti_csc ti_vpdma v4l2_mem2mem v4l2_common videobuf2_memops videobuf2_v4l2 videobuf2_core videodev media btsdio bluetooth joydev ecdh_generic brcmfmac stmpe_adc brcmutil cfg80211 omap_remoteproc virtio_rpmsg_bus rpmsg_core usb_f_ncm usb_f_mass_storage usb_f_rndis u_ether libcomposite uio_pdrv_genirq uio cmemk(O)
[ 889.236083] CPU: 0 PID: 1810 Comm: bash Tainted: G W O 4.14.108-ti-r137 #1buster
[ 889.236090] Hardware name: Generic DRA74X (Flattened Device Tree)
[ 889.236111] [] (unwind_backtrace) from [] (show_stack+0x20/0x24)
[ 889.236125] [] (show_stack) from [] (dump_stack+0x80/0x94)
[ 889.236139] [] (dump_stack) from [] (__warn+0xec/0x114)
[ 889.236151] [] (__warn) from [] (warn_slowpath_fmt+0x58/0x74)
[ 889.236164] [] (warn_slowpath_fmt) from [] (l3_interrupt_handler+0x368/0x3a4)
[ 889.236180] [] (l3_interrupt_handler) from [] (__handle_irq_event_percpu+0x84/0x2d0)
[ 889.236193] [] (__handle_irq_event_percpu) from [] (handle_irq_event_percpu+0x3c/0x90)
[ 889.236204] [] (handle_irq_event_percpu) from [] (handle_irq_event+0x48/0x6c)
[ 889.236216] [] (handle_irq_event) from [] (handle_fasteoi_irq+0xd0/0x178)
[ 889.236229] [] (handle_fasteoi_irq) from [] (generic_handle_irq+0x34/0x44)
[ 889.236240] [] (generic_handle_irq) from [] (__handle_domain_irq+0x8c/0xfc)
[ 889.236252] [] (__handle_domain_irq) from [] (gic_handle_irq+0x4c/0x88)
[ 889.236264] [] (gic_handle_irq) from [] (__irq_svc+0x6c/0xa8)
[ 889.236271] Exception stack(0xce9bbc28 to 0xce9bbc70)
[ 889.236282] bc20: 0000000a ce9bbcac 00000025 c10417a6 ce9bbcac c104179c
[ 889.236292] bc40: ce9bbd0c d3761420 c108c08c ce9bbdac ce9bbcac ce9bbc8c ce9bbc90 ce9bbc78
[ 889.236300] bc60: c0cc6b40 c0cc2dac 20010013 ffffffff
[ 889.236314] [] (__irq_svc) from [] (format_decode+0x78/0x5e8)
[ 889.236326] [] (format_decode) from [] (vsnprintf+0x84/0x3f4)
[ 889.236338] [] (vsnprintf) from [] (snprintf+0x38/0x54)
[ 889.236353] [] (snprintf) from [] (dev_vprintk_emit+0x64/0x210)
[ 889.236367] [] (dev_vprintk_emit) from [] (dev_printk_emit+0x40/0x5c)
[ 889.236379] [] (dev_printk_emit) from [] (__dev_printk+0x58/0x98)
[ 889.236391] [] (__dev_printk) from [] (_dev_info+0x58/0x74)
[ 889.236406] [] (_dev_info) from [] (rproc_boot+0x574/0x634)
[ 889.236421] [] (rproc_boot) from [] (state_store+0x80/0x144)
[ 889.236435] [] (state_store) from [] (dev_attr_store+0x28/0x34)
[ 889.236450] [] (dev_attr_store) from [] (sysfs_kf_write+0x48/0x54)
[ 889.236463] [] (sysfs_kf_write) from [] (kernfs_fop_write+0xf8/0x1dc)
[ 889.236477] [] (kernfs_fop_write) from [] (__vfs_write+0x28/0x48)
[ 889.236490] [] (__vfs_write) from [] (vfs_write+0xb4/0x1c8)
[ 889.236503] [] (vfs_write) from [] (SyS_write+0x5c/0xbc)
[ 889.236516] [] (SyS_write) from [] (ret_fast_syscall+0x0/0x54)
[ 889.236524] —[ end trace 0de72da1c33b76f5 ]—
[ 889.236536] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 1267.955180] pru-rproc 4b2b4000.pru: can’t change firmware while running
[ 1271.783200] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 1281.595190] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 1281.595471] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 1281.595490] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 1284.647130] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 2370.791112] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 2370.791406] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 2370.791426] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 2376.151149] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 2380.387163] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 2380.387284] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 2380.387303] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 2382.287146] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 2393.327159] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 2393.327706] remoteproc remoteproc6: Booting fw image PRU_ucfsm.out, size 370028
[ 2393.327762] ------------[ cut here ]------------
[ 2393.327782] WARNING: CPU: 0 PID: 1810 at drivers/bus/omap_l3_noc.c:147 l3_interrupt_handler+0x368/0x3a4
[ 2393.327790] 44000000.ocp:L3 Standard Error: MASTER PRUSS2 PRU1 TARGET PCIE1 (Read): At Address: 0x00806664 : Data Access in Supervisor mode during Functional access
[ 2393.327796] Modules linked in: rpmsg_rpc rpmsg_proto pru_rproc pruss_intc bnep pruss pvrsrvkm(O) pruss_soc_bus ti_vpe ti_sc videobuf2_dma_contig ti_csc ti_vpdma v4l2_mem2mem v4l2_common videobuf2_memops videobuf2_v4l2 videobuf2_core videodev media btsdio bluetooth joydev ecdh_generic brcmfmac stmpe_adc brcmutil cfg80211 omap_remoteproc virtio_rpmsg_bus rpmsg_core usb_f_ncm usb_f_mass_storage usb_f_rndis u_ether libcomposite uio_pdrv_genirq uio cmemk(O)
[ 2393.327996] CPU: 0 PID: 1810 Comm: bash Tainted: G W O 4.14.108-ti-r137 #1buster
[ 2393.328002] Hardware name: Generic DRA74X (Flattened Device Tree)
[ 2393.328024] [] (unwind_backtrace) from [] (show_stack+0x20/0x24)
[ 2393.328039] [] (show_stack) from [] (dump_stack+0x80/0x94)
[ 2393.328052] [] (dump_stack) from [] (__warn+0xec/0x114)
[ 2393.328064] [] (__warn) from [] (warn_slowpath_fmt+0x58/0x74)
[ 2393.328078] [] (warn_slowpath_fmt) from [] (l3_interrupt_handler+0x368/0x3a4)
[ 2393.328093] [] (l3_interrupt_handler) from [] (__handle_irq_event_percpu+0x84/0x2d0)
[ 2393.328107] [] (__handle_irq_event_percpu) from [] (handle_irq_event_percpu+0x3c/0x90)
[ 2393.328118] [] (handle_irq_event_percpu) from [] (handle_irq_event+0x48/0x6c)
[ 2393.328130] [] (handle_irq_event) from [] (handle_fasteoi_irq+0xd0/0x178)
[ 2393.328143] [] (handle_fasteoi_irq) from [] (generic_handle_irq+0x34/0x44)
[ 2393.328154] [] (generic_handle_irq) from [] (__handle_domain_irq+0x8c/0xfc)
[ 2393.328166] [] (__handle_domain_irq) from [] (gic_handle_irq+0x4c/0x88)
[ 2393.328178] [] (gic_handle_irq) from [] (__irq_svc+0x6c/0xa8)
[ 2393.328185] Exception stack(0xce9bbc20 to 0xce9bbc68)
[ 2393.328196] bc20: c104179c ce9bbcac 00000025 c10417a6 ce9bbcac c104179c ce9bbd0c d3761420
[ 2393.328206] bc40: c108c08c ce9bbdac ce9bbcac ce9bbc8c ce9bbc90 ce9bbc70 c0cc6b40 c0cc2da4
[ 2393.328214] bc60: 20010013 ffffffff
[ 2393.328227] [] (__irq_svc) from [] (format_decode+0x70/0x5e8)
[ 2393.328239] [] (format_decode) from [] (vsnprintf+0x84/0x3f4)
[ 2393.328252] [] (vsnprintf) from [] (snprintf+0x38/0x54)
[ 2393.328267] [] (snprintf) from [] (dev_vprintk_emit+0x64/0x210)
[ 2393.328280] [] (dev_vprintk_emit) from [] (dev_printk_emit+0x40/0x5c)
[ 2393.328293] [] (dev_printk_emit) from [] (__dev_printk+0x58/0x98)
[ 2393.328305] [] (__dev_printk) from [] (_dev_info+0x58/0x74)
[ 2393.328320] [] (_dev_info) from [] (rproc_boot+0x574/0x634)
[ 2393.328335] [] (rproc_boot) from [] (state_store+0x80/0x144)
[ 2393.328348] [] (state_store) from [] (dev_attr_store+0x28/0x34)
[ 2393.328363] [] (dev_attr_store) from [] (sysfs_kf_write+0x48/0x54)
[ 2393.328377] [] (sysfs_kf_write) from [] (kernfs_fop_write+0xf8/0x1dc)
[ 2393.328391] [] (kernfs_fop_write) from [] (__vfs_write+0x28/0x48)
[ 2393.328404] [] (__vfs_write) from [] (vfs_write+0xb4/0x1c8)
[ 2393.328416] [] (vfs_write) from [] (SyS_write+0x5c/0xbc)
[ 2393.328431] [] (SyS_write) from [] (ret_fast_syscall+0x0/0x54)
[ 2393.328439] —[ end trace 0de72da1c33b76f6 ]—
[ 2393.328452] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 2397.083151] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru
[ 2406.111125] remoteproc remoteproc6: powering up 4b2b4000.pru
[ 2406.111417] remoteproc remoteproc6: Booting fw image PRU-LEDs.out, size 32644
[ 2406.111437] remoteproc remoteproc6: remote processor 4b2b4000.pru is now up
[ 2408.331136] remoteproc remoteproc6: stopped remote processor 4b2b4000.pru

2

Looks like PRU attempts to access PCIE1 address space. I suspect it’s a bug in your PRU firmware.

MASTER PRUSS2 PRU1 TARGET PCIE1 (Read): At Address: 0x00806664

TI has tutorials how to use JTAG to debug PRU. Another option is https://markayoder.github.io/PRUCookbook/04debug/debug.html

Regards,
Dimitar

3

Hi Dimitar,

Thanks for your reply, yes, I don’t understand that as its code to flash lights. Its built with different states, which makes it more complicated, but only flashes LEDs.

How can I know what the kernel module is doing so I can see more details and know where to look? Such as, is this happening when the firmware is being processed by the module, or is this the module giving the error once the code is trying to run on the PRU?

Thanks,
Fisher

4

The error message is emitted from the system bus driver (drivers/bus/omap_l3_noc.c ).

I interpret it as a bug in your PRU firmware. When issue occurs, please try to inspect the PRU state. See https://zeekhuge.me/post/ptp_docs_commands_and_tools/ , or use JTAG.

Regards,
Dimitar

5

Hi all,

I solved my issue in the end of it not booting, or causing errors with “L3_main” in the title from remoteProc to dmesg.

The issue turned out to be both having interrupts from the IEP timer, and that C++ needs .init_array in the linker command file to actually initialise the constructors.
The linker did add the .init_array section at the end of the binary, but I don’t think actually got the code to jump to the needed addresses without that section specifically put into the linker file.

I assume the interrupt issue may be been due to using an empty resource_table, so the kernel module didn’t set anything up to receive them on the Linux side, but then the timer sent them.
I’m not fully sure about that because I believe I’ve also left them enabled, and not had dmesg errors.

I still can’t find any RemoteProc documentation of the possible config options, or how to get extra info out of the kernel modules, such as what it’s doing as it processes the resource_table.

Fisher

6

Hello Fischer

This file looks like it’s processing the resource tablehttps://docs.huihoo.com/doxygen/linux/kernel/3.7/remoteproc__core_8c_source.html

7

Hi Mark,

Thanks for your reply, yes, I’ve not seen a lot posted about remoteProc when I searched, and I know it has to be an important issue due to at least PRU interest being almost totally based on it. Even with JTAG and directly programming your code into the PRUs, Linux still has to receive the interrupts.
Plus, as you mentioned, more chips are multi-architecture. We’re spoilt for choice with dual core M4 and C6000 DSP in the AI, and they need to be used properly to take advantage of this chip.
I think too many people likely use it like a Raspberry Pi which defeats the point, as the Pi is multi-core Cortex A just for Linux, while the BeagleBone AI and Black have less Linux CPU power, but excel in using the PRUs and M3/M4 etc.

Thanks a lot for the documentation link, I’ve only more recently returned to programming, so have had to refresh myself in not just C and C++, but also micro controller architecture and compile options with Makefiles.

Ya, I’ll look over the documentation you sent a little later, thankfully that documentation framework exists, as documentation is the last thing programmers want to do.

I’m currently sorting out the IEP timer use so a simple real-time scheduler can run using it. It will run state-machines.

I’ll document all I’ve done in the end, my supervisor should want to go through and replicate what I’ve done to confirm the steps are correct, and then people will have more of a guide.

Fisher

8

Hi Fischer

what I meant to say is the AM57xx is definitely more complicated as there’s more cores and that will definitely break quickly without correctly modifying the table for the DSP or M4.

There’s absolutely no docs beyond the SDK and many people are not using SDK they use the Debian build wiki so I think that’s a weak point in getting users to use x15 or AI.

Here’s the link for thae AM57x Linux IPC

If you dig around in this document you see some messages being displayed for normal loads and rpmsg and it discusses changing resource table.

Also look at the details involved in getting complicated examples working in CCS.

Again the Am35 build may be set up to handle most PRU examples and while I could envision debugging it’s PRUs with printf I can’t imagine trying complex DSP or a M4 without CCS and JTAG to debug for exactly the reasons you point out the documents are scarce. Buts just me.

Here’s the 57x document as you can see in it the DSP and M4 have MMU on the interconnect between the ARM Host.

http://software-dl.ti.com/processor-sdk-linux/esd/docs/latest/linux/Foundational_Components_IPC.html#ipc-for-am57xx

If anything this overview should give you more confidence deciphering what your original question was and still remains.

My understanding is hazy but the carvout and increasing of any shared memory sizes is where you need to change this table and I’ve seen no other documents discuss this for.

I suspect the existing working PRU on Am35x examples are very simple in nature as you said in last post simple demos work. You mentioned using interrupts that’s probably a faster way to get Data to the Host. To me what good is a determistic fast IO processor if the Data needs to be acted on quickly as in a hard realtime control theory loop.
The other cores the DSP and M4 have edma also on the interconnect not sure about PRU. It’s all described in the document below.

Your simple example comment bthat’s what caught my eye I think your right :thinking:. Most people use the PRU for dedicated offloading I doubt many use every function available for IO like a normal processor application I might be wrong but I do know the DSP and M4 share many peripheral with the ARM so full blown applications on those cores can and will be designed eventually.
I know this as I have seen product that’s in field doing so. How those engineer’s deciphering questions like yours I don’t know I suspect they had dedicated engineer’s at TI assigned to the company as they were buying large quantities of chips then and in the past.

With open source it’s kind of like the book I’m reading about Google and Stanford students. The professor assigns summer projects to their PHD candidate’s :crazy_face::face_with_hand_over_mouth: and in the case of the Google Co-founders they crashed the University computer and we’re nicely ask to leave the campus but told they could come back anytime if Google didn’t work out.
And they paid the tuition.

Sorry for detour there but I think this is were Google’s Summer projects are inspired from Stanford.

Keep us updated if you find more details

Mark