Dear support,
Our goal is to use the C6670 as part as a full system. In this system, a host card (single board computer with an OS) will be loading the code inside the DSP and the DSP will then be running. We will have in the system 8 cards all connected via the PCIe bus.
That being said, we want, after power on to the system, that the C6670 would be ready to receive a new program.
It seemed logical to study how c:\TI_MCSDK\mcsdk_2_01_00_03\tools\boot_loader\examples\pcie\ was running under linux to learn and replicate the process using another os. I am not an expert in linux but we kind of got the examples working.
We are having some issues that I need your help with. the first part are purely linux and the other ones are more about the loading system on boot.
We have EEPROM 0x51 loaded with IBL from mcsdk_2_01_00_03 without recompiling the code. the binary provided is the one in flash.
----------------------- linux questions ----------------
1) let's say that I compile the pciedemo to run hello_word. then, I insert the module into the kernel, all seems to be ok. When I remove the module, recompile and reinssert the module, the linux PC completely hang and crash. It needs a power off to be able to run again. This means I need to power off each time I want to try my code. Nothing in the instructions says that I cannot do insert, remove and insert again. Should I do something in between 2 insertions? Is there anyway I can provide you more information that can help with this issue?
we run a fedora distribution with this version of the kernel: 2.6.35.6-45.fc14.x86_64
2) I searched for this problem on the forum and no one seems to have ask about it so I hope it is a linux issue. when trying to run the pcieInterrupt example to use EDMA, I receive a segmentation fault. I am attaching a file
[root@localhost spectrum]# lspci -d 104c:b005 -xxx -v 08:00.0 Multimedia controller: Texas Instruments Device b005 (rev 01) Physical Slot: 4 Flags: bus master, fast devsel, latency 0, IRQ 16 Memory at e4400000 (32-bit, non-prefetchable) [size=1M] Memory at e4200000 (32-bit, prefetchable) [size=1M] Memory at e4000000 (32-bit, prefetchable) [size=2M] Memory at e3000000 (32-bit, prefetchable) [size=16M] Memory at e4310000 (32-bit, prefetchable) [size=4K] Memory at e4300000 (32-bit, prefetchable) [size=64K] Capabilities: [40] Power Management version 3 Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit+ Capabilities: [70] Express Endpoint, MSI 00 Capabilities: [100] Advanced Error Reporting Kernel modules: windrvr6 00: 4c 10 05 b0 07 01 18 00 01 00 80 04 10 00 00 00 10: 00 00 40 e4 08 00 20 e4 08 00 00 e4 08 00 00 e3 20: 08 00 31 e4 08 00 30 e4 00 00 00 00 00 00 01 00 30: 00 00 00 00 40 00 00 00 00 00 00 00 0a 01 00 00 40: 01 50 03 00 00 00 00 00 00 00 00 00 00 00 00 00 50: 05 70 80 00 00 00 00 00 00 00 00 00 00 00 00 00 60: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 70: 10 00 02 00 01 87 2c 01 1f 28 01 00 22 34 03 00 80: c0 00 22 10 00 00 00 00 00 00 00 00 00 00 00 00 90: 00 00 00 00 1f 00 00 00 00 00 00 00 06 00 00 00 a0: 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 c0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fedora kernel version 2.6.35.6-45.fc14.x86_64 [ 1641.392501] Finding the device.... [ 1641.392517] Found TI device [ 1641.392519] TI device: vendor=0x104c, dev=0xb005, irq=0x0000000a [ 1641.392520] Reading the BAR areas.... [ 1641.394245] Enabling the device.... [ 1641.394259] pci 0000:08:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [ 1641.394263] pci 0000:08:00.0: setting latency timer to 64 [ 1641.394270] Access PCIE application register .... [ 1641.394272] Registering the irq 10 ... [ 1641.394288] Allocating consistent memory ... [ 1641.402129] Boot entry address is 0x 82ee80 [ 1641.402743] Total 5 sections, 0xfed0 bytes of data were written [ 1643.462014] Write DMA to DSP ... [ 1643.468102] Generating interrupt to DSP ... [ 1643.925097] irq 16: nobody cared (try booting with the "irqpoll" option) [ 1643.925100] Pid: 2653, comm: insmod Tainted: P I 2.6.35.6-45.fc14.x86_64 #1 [ 1643.925102] Call Trace: [ 1643.925103] <IRQ> [<ffffffff810a6e2b>] __report_bad_irq.clone.1+0x3d/0x8b [ 1643.925111] [<ffffffff810a6f93>] note_interrupt+0x11a/0x17f [ 1643.925113] [<ffffffff810a7a73>] handle_fasteoi_irq+0xa8/0xce [ 1643.925116] [<ffffffff8100c2ea>] handle_irq+0x88/0x90 [ 1643.925120] [<ffffffff8146efb4>] do_IRQ+0x5c/0xb4 [ 1643.925122] [<ffffffff81469513>] ret_from_intr+0x0/0x11 [ 1643.925124] <EOI> [<ffffffff81010b17>] ? native_read_tsc+0x6/0x16 [ 1643.925129] [<ffffffff8122080e>] paravirt_read_tsc+0xe/0x12 [ 1643.925131] [<ffffffff812208ff>] delay_tsc+0x35/0x74 [ 1643.925133] [<ffffffff81220859>] __delay+0xf/0x11 [ 1643.925135] [<ffffffff8122089d>] __const_udelay+0x42/0x44 [ 1643.925139] [<ffffffffa00f3386>] init_module+0x324/0x424 [pciedemo] [ 1643.925141] [<ffffffff81010587>] ? sched_clock+0x9/0xd [ 1643.925144] [<ffffffff8103c040>] ? need_resched+0x23/0x2d [ 1643.925146] [<ffffffff8103c058>] ? should_resched+0xe/0x2e [ 1643.925149] [<ffffffff81467b05>] ? _cond_resched+0xe/0x22 [ 1643.925150] [<ffffffff8146824a>] ? mutex_lock+0x29/0x50 [ 1643.925153] [<ffffffff810c2f5c>] ? trace_module_notify+0x2b5/0x2c6 [ 1643.925156] [<ffffffff810ad93a>] ? tracepoint_module_notify+0x2c/0x30 [ 1643.925158] [<ffffffff8146c4b4>] ? notifier_call_chain+0x37/0x63 [ 1643.925162] [<ffffffffa00f3062>] ? init_module+0x0/0x424 [pciedemo] [ 1643.925165] [<ffffffff810021a1>] do_one_initcall+0x5e/0x155 [ 1643.925168] [<ffffffff8107caa9>] sys_init_module+0xa6/0x1e4 [ 1643.925170] [<ffffffff81009cf2>] system_call_fastpath+0x16/0x1b [ 1643.925172] handlers: [ 1643.925173] [<ffffffffa009160b>] (nouveau_irq_handler+0x0/0x1ac5 [nouveau]) [ 1643.925187] Disabling IRQ #16 [ 1644.467454] DMA write throughput is: 657.24 MB/s [ 1644.467470] divide error: 0000 [#1] SMP [ 1644.467473] last sysfs file: /sys/devices/system/cpu/cpu1/cache/index2/shared_cpu_map [ 1644.467475] CPU 1 [ 1644.467476] Modules linked in: pciedemo(+) fuse sunrpc ipv6 cpufreq_ondemand acpi_cpufreq freq_table mperf uinput snd_hda_codec_realtek windrvr6(P) hp_wmi rfkill snd_hda_intel x38_edac edac_core iTCO_wdt iTCO_vendor_support snd_hda_codec snd_hwdep ppdev parport_pc parport wmi snd_seq snd_seq_device tg3 snd_pcm snd_timer snd soundcore snd_page_alloc microcode nouveau ttm drm_kms_helper drm i2c_algo_bit video output i2c_core [last unloaded: scsi_wait_scan] [ 1644.467499] [ 1644.467501] Pid: 2653, comm: insmod Tainted: P I 2.6.35.6-45.fc14.x86_64 #1 0AA0h/HP xw4600 Workstation [ 1644.467503] RIP: 0010:[<ffffffffa00f3404>] [<ffffffffa00f3404>] init_module+0x3a2/0x424 [pciedemo] [ 1644.467507] RSP: 0018:ffff88005e39bde8 EFLAGS: 00010246 [ 1644.467509] RAX: 00000000003d0900 RBX: 00000000003d0900 RCX: 0000000000000000 [ 1644.467511] RDX: 0000000000000000 RSI: 0000000000000096 RDI: ffffffffa00f38e3 [ 1644.467512] RBP: ffff88005e39bf08 R08: 0000000000000002 R09: 00000000fffffffe [ 1644.467514] R10: ffff8800de39bd07 R11: 0000000000000000 R12: 0000000000000340 [ 1644.467516] R13: 00007f28bc572010 R14: 00007f28bc572010 R15: 0000000000000003 [ 1644.467518] FS: 00007f28bc634720(0000) GS:ffff880002080000(0000) knlGS:0000000000000000 [ 1644.467520] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1644.467521] CR2: 00007f48e8605fc8 CR3: 000000005e3c5000 CR4: 00000000000406e0 [ 1644.467523] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1644.467525] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1644.467527] Process insmod (pid: 2653, threadinfo ffff88005e39a000, task ffff88007b30c5c0) [ 1644.467528] Stack: [ 1644.467529] ffff880000000000 ffffffff81010587 ffff88005e39be08 ffffffff8103c040 [ 1644.467532] <0> ffff88005e39be18 ffffffff8103c058 ffff88005e39be28 ffffffff81467b05 [ 1644.467534] <0> ffff88005e39be58 ffffffff8146824a ffff88005e39be48 0000000000000246 [ 1644.467537] Call Trace: [ 1644.467540] [<ffffffff81010587>] ? sched_clock+0x9/0xd [ 1644.467543] [<ffffffff8103c040>] ? need_resched+0x23/0x2d [ 1644.467545] [<ffffffff8103c058>] ? should_resched+0xe/0x2e [ 1644.467547] [<ffffffff81467b05>] ? _cond_resched+0xe/0x22 [ 1644.467549] [<ffffffff8146824a>] ? mutex_lock+0x29/0x50 [ 1644.467552] [<ffffffff810c2f5c>] ? trace_module_notify+0x2b5/0x2c6 [ 1644.467554] [<ffffffff810ad93a>] ? tracepoint_module_notify+0x2c/0x30 [ 1644.467556] [<ffffffff8146c4b4>] ? notifier_call_chain+0x37/0x63 [ 1644.467560] [<ffffffffa00f3062>] ? init_module+0x0/0x424 [pciedemo] [ 1644.467562] [<ffffffff810021a1>] do_one_initcall+0x5e/0x155 [ 1644.467565] [<ffffffff8107caa9>] sys_init_module+0xa6/0x1e4 [ 1644.467567] [<ffffffff81009cf2>] system_call_fastpath+0x16/0x1b [ 1644.467568] Code: 33 08 01 00 48 2b 05 1c 08 01 00 31 d2 48 8b 0d 2b 08 01 00 2b 0d 15 08 01 00 48 c7 c7 e3 38 0f a0 69 c0 40 42 0f 00 01 c1 89 d8 <f7> f1 89 c6 6b c2 64 31 d2 f7 f1 89 c2 31 c0 e8 05 3f 37 e1 48 [ 1644.467590] RIP [<ffffffffa00f3404>] init_module+0x3a2/0x424 [pciedemo] [ 1644.467593] RSP <ffff88005e39bde8> [ 1644.467596] ---[ end trace a7919e7f17c0a727 ]--- [root@localhost spectrum]# lsmod Module Size Used by tcp_lp 2111 0 pciedemo 77810 1
From this dump, can you help me to find out what the problem is?
I think there is something wrong with the IRQ. the lspci says that the IRQ is 16 and when enabling the devicem we see this printout. pci 0000:08:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16, however, in the following of dmesg log, irq10 is being registered (Registering the irq 10)
----------------------- other questions ----------------
3) Why can't a *.out file be loaded directly onto memory? When loading from CCS, is CCS studio going through the same process of converting the *.out file into a boot table file
4) Is it required t be in PCIe boot mode to run the examples provided? If I boot from flash with the PCIe module enabled, and the code from flash do similar things that the current IBL, would the examples work?
5) Are you aware of any power on issues using PCIe boot mode? The reason for this question is that when plugging the EVM in different PC's the EVM is not working properly. Some PC needs a warm reset before the OS starts, some PC do not start at all, the displays is getting completely corrupted, Some PC's do three strange on-off sequence before the PC boots.
Thanks for your help
Aymeric
