June 2014 - U-Boot

Re: [U-Boot] What file formats can be loaded and launched by u-boot
by Wolfgang Denk 11 Jun '14

11 Jun '14

Dear "Misbah", In message <005301cf8578$18686be0$493943a0$@khan@tes-dst.com> you wrote: > > I have an RTOS compiled for ARM cotex-A9 generating .out file format. ".out" is not any known linker output format. Do you mean a.out? It has been ages since I've seen this used the last time - way back in the previous millenium. > How can I use u-boot.bin to download the RTOS image from an SD card "FAT32 > formatted" and launch the RTOS kernel In general, you can use objcopy (from the GNU binutils package) to convert into raw binary format. This you can then either load directly, or wrap it using the mkimage tool into either a legacy U-Boot image or a FIT image, which then can be started by U-Boot. Best regards, Wolfgang Denk -- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd(a)denx.de : 1. What is the possibility of this being added in the future? In the near future, the probability is close to zero. In the distant future, I'll be dead, and posterity can do whatever they like... :-) - lwall

1 0

[U-Boot] What file formats can be loaded and launched by u-boot
by Misbah 11 Jun '14

11 Jun '14

Hello All, I have an RTOS compiled for ARM cotex-A9 generating .out file format. How can I use u-boot.bin to download the RTOS image from an SD card "FAT32 formatted" and launch the RTOS kernel Can you please share your views. Regards, Misbah

1 0

[U-Boot] USB EHCI problem on NeufBox V6
by Adrien Ricciardi 11 Jun '14

11 Jun '14

Hello, I am currently working on a Broadcom BCM63168 SoC board (NeufBox V6). This is a MIPS32 big endian CPU. My goal is to replace the CFERAM stage by U-Boot 2014.01 (I keep the CFEROM for now as it contains a lot of magic). I managed to make U-Boot start on the platform. It can mount an UBIFS file system stored on the NAND chip and load Linux. Now, I would like to add the USB Host support to U-Boot in order to boot from an USB stick. Thus, I tried to make EHCI drivers based on the working Linux ones. The SoC integrates a (supposed) standard EHCI controller. However, I can't start the USB subsystem (using the command "usb start") because the USB enumeration process crashes when it sends a GetDescriptor() request to every device but the builtin root hub. When the U-Boot USB calls ehci_submit_async(), the EHCI controller times out (no matter what the timeout delay is). When the ehci_submit_async() function is called another time, a bad cache address (0x00000000) is provided to the function : /* * Invalidate the memory area occupied by buffer * Don't try to fix the buffer alignment, if it isn't properly * aligned it's upper layer's fault so let invalidate_dcache_range() * vow about it. But we have to fix the length as it's actual * transfer length and can be unaligned. This is potentially * dangerous operation, it's responsibility of the calling * code to make sure enough space is reserved. */ invalidate_dcache_range((uint32_t)buffer, ALIGN((uint32_t)buffer + length, ARCH_DMA_MINALIGN)); and everything crashes. By disconnecting every USB hub and device connected to the SoC USB Host peripheral, only the builtin root hub is recognized by the command "usb start" and U-Boot does not crash. I tried to connect an USB sniffer on the bus (an Ellisys 200) to capture the USB frames. The EHCI timeout occurs when the USB subsystem tries to send a GetDescriptor() request (using ehci_submit_async()). Only bad Start Of Frames are captured by the USB sniffer, and when U-Boot crashes the bad SOF continue to be emitted (I think the EHCI controller is hanged too). I tried to update the drivers/usb U-Boot directory with the one from U-Boot 2014.04 and with the one from the git repository, but the problem could not be solved. Have you got any idea on what can prevent the EHCI controller to communicate with the external USB bus through the SoC root hub ? Best regards, Adrien RICCIARDI

1 0

[U-Boot] AMP implementation
by Misbah 11 Jun '14

11 Jun '14

Hello All, I have a requirement to have AMP on an multi core processor (i.MX6 freescale) using linux as OS The uboot should download uImage_core0 and uImage_core1 and be able to launch both the kernel Running on core0 and core1 of the processor. We need to make necessary changes to the uboot such that bootm command could be used accordingly Can you help me to achieve this ? With your valuable sugessions ? Thank You, Regards, Misbah

1 0

[U-Boot] AMP of linux on multi core processor and u-boot
by Misbah 11 Jun '14

11 Jun '14

Hello All, I have a requirement to have AMP on an multi core processor (i.MX6 freescale) using linux as OS The uboot should download uImage_core0 and uImage_core1 and be able to launch both the kernel Running on core0 and core1 of the processor. We need to make necessary changes to the uboot such that bootm command could be used accordingly Can you help me to achieve this ? With your valuable sugessions ? Thank You, Regards, Misbah

1 0

[U-Boot] [U-boot] sandbox test script question
by TigerLiu＠via-alliance.com 11 Jun '14

11 Jun '14

Hi, Simon: I tried to run test/image scripts: test/image/test-imagetools.sh - multi-file images test/image/test-fit.py - FIT images 1. test-imagetools.sh need to copy u-boot/tools directory to sandbox dir manually, otherwise it would fail. 2. test-fit.py it failed. Error msgs: IOError : [Errno 2] No such file or directory: '/tmp/tmpfLQZPq/kernel-out.bin' And, run 'ls /tmp/tmpfLQZPq' It had test-kernel.bin, not kernel-out.bin Best wishes,

3 4

[U-Boot] [PULL] u-boot-usb/master
by Marek Vasut 11 Jun '14

11 Jun '14

The following changes since commit 90b51c33f362926e17d4c07dcef1ce822abaa89f: Merge branch 'master' of git://git.denx.de/u-boot-arm (2014-05-30 11:34:39 -0400) are available in the git repository at: git://git.denx.de/u-boot-usb.git master for you to fetch changes up to 006c7026882011ba49c9a39d27c4aff3ace07847: usb: ci_udc: complete ep0 direction handling (2014-06-01 19:22:40 +0200) ---------------------------------------------------------------- Lukasz Majewski (1): dfu: Introduction of the "dfu_hash_algo" env variable for checksum method setting Stephen Warren (6): usb: hub: fix power good delay timing usb: hub: remove CONFIG_USB_HUB_MIN_POWER_ON_DELAY usb: ci_udc: detect queued requests on ep0 usb: ci_udc: use a single descriptor for ep0 usb: ci_udc: pre-allocate ep0 req usb: ci_udc: complete ep0 direction handling README | 3 --- common/usb_hub.c | 11 +++++------ drivers/dfu/dfu.c | 52 +++++++++++++++++++++++++++++++++++++++++++++++----- drivers/usb/gadget/ci_udc.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++----------------- drivers/usb/gadget/ci_udc.h | 2 ++ include/configs/cm_t35.h | 2 -- include/configs/gw_ventana.h | 1 - 7 files changed, 128 insertions(+), 34 deletions(-)

2 4

[U-Boot] [PATCH] dfu: Disable default calculation of CRC32
by Lukasz Majewski 11 Jun '14

11 Jun '14

Patch (SHA1: bd694244db7bc969954) dfu: Introduction of the "dfu_hash_algo" env variable for checksum method setting already introduced more generic handling of the crc32 calculation. Up till now the CRC32 of received data was calculated unconditionally. This patch changes this and from now - by default the crc32 is NOT calculated anymore. Signed-off-by: Lukasz Majewski <l.majewski(a)samsung.com> Cc: Marek Vasut <marex(a)denx.de> --- drivers/dfu/dfu.c | 20 +++++++------------- 1 file changed, 7 insertions(+), 13 deletions(-) diff --git a/drivers/dfu/dfu.c b/drivers/dfu/dfu.c index 5878f99..dc09ff6 100644 --- a/drivers/dfu/dfu.c +++ b/drivers/dfu/dfu.c @@ -106,21 +106,15 @@ static char *dfu_get_hash_algo(void) char *s; s = getenv("dfu_hash_algo"); - /* - * By default the legacy behaviour to calculate the crc32 hash - * value is preserved. - * - * To disable calculation of the hash algorithm for received data - * specify the "dfu_hash_algo = disabled" at your board envs. - */ - debug("%s: DFU hash method: %s\n", __func__, s ? s : "not specified"); - - if (!s || !strcmp(s, "crc32")) - return "crc32"; - - if (!strcmp(s, "disabled")) + if (!s) return NULL; + if (!strcmp(s, "crc32")) { + debug("%s: DFU hash method: %s\n", __func__, s); + return s; + } + + error("DFU hash method: %s not supported!\n", s); return NULL; } -- 2.0.0.rc2

2 1

[U-Boot] [PATCH] usb: ci_udc: terminate ep0 INs with a zlp when required
by Stephen Warren 11 Jun '14

11 Jun '14

From: Stephen Warren <swarren(a)nvidia.com> Sometimes, a zero-length packet is required at the end of an IN transaction so that the host knows the device is done sending data. Enhance ci_udc to send a zlp when necessary. See the comments for more details. Signed-off-by: Stephen Warren <swarren(a)nvidia.com> --- This depends on all the other ci_udc patches I've sent previously at least for diff context if nothing else. --- drivers/usb/gadget/ci_udc.c | 38 +++++++++++++++++++++++++++++++++++--- 1 file changed, 35 insertions(+), 3 deletions(-) diff --git a/drivers/usb/gadget/ci_udc.c b/drivers/usb/gadget/ci_udc.c index 435a2720e96f..b18bee43ad89 100644 --- a/drivers/usb/gadget/ci_udc.c +++ b/drivers/usb/gadget/ci_udc.c @@ -369,18 +369,50 @@ static void ci_ep_submit_next_request(struct ci_ep *ci_ep) ci_req = list_first_entry(&ci_ep->queue, struct ci_req, queue); len = ci_req->req.length; - item->next = TERMINATE; - item->info = INFO_BYTES(len) | INFO_IOC | INFO_ACTIVE; + item->info = INFO_BYTES(len) | INFO_ACTIVE; item->page0 = (uint32_t)ci_req->hw_buf; item->page1 = ((uint32_t)ci_req->hw_buf & 0xfffff000) + 0x1000; item->page2 = ((uint32_t)ci_req->hw_buf & 0xfffff000) + 0x2000; item->page3 = ((uint32_t)ci_req->hw_buf & 0xfffff000) + 0x3000; item->page4 = ((uint32_t)ci_req->hw_buf & 0xfffff000) + 0x4000; - ci_flush_qtd(num); head->next = (unsigned) item; head->info = 0; + /* + * When sending the data for an IN transaction, the attached host + * knows that all data for the IN is sent when one of the following + * occurs: + * a) A zero-length packet is transmitted. + * b) A packet with length that isn't an exact multiple of the ep's + * maxpacket is transmitted. + * c) Enough data is sent to exactly fill the host's maximum expected + * IN transaction size. + * + * One of these conditions MUST apply at the end of an IN transaction, + * or the transaction will not be considered complete by the host. If + * none of (a)..(c) already applies, then we must force (a) to apply + * by explicitly sending an extra zero-length packet. + */ + /* IN !a !b !c */ + if (in && len && !(len % ci_ep->ep.maxpacket) && ci_req->req.zero) { + /* + * Each endpoint has 2 items allocated, even though typically + * only 1 is used at a time since either an IN or an OUT but + * not both is queued. For an IN transaction, item currently + * points at the second of these items, so we know that we + * can use (item - 1) to transmit the extra zero-length packet + */ + item->next = (unsigned)(item - 1); + item--; + item->info = INFO_ACTIVE; + } + + item->next = TERMINATE; + item->info |= INFO_IOC; + + ci_flush_qtd(num); + DBG("ept%d %s queue len %x, req %p, buffer %p\n", num, in ? "in" : "out", len, ci_req, ci_req->hw_buf); ci_flush_qh(num); -- 1.8.1.5

2 1

[U-Boot] [PATCH 1/4] usb: ci_udc: call udc_disconnect() from ci_pullup()
by Stephen Warren 11 Jun '14

11 Jun '14

From: Stephen Warren <swarren(a)nvidia.com> ci_pullup()'s !is_on path contains a cut/paste copy of udc_disconnect(). Remove the duplication by simply calling udc_disconnect() instead. Signed-off-by: Stephen Warren <swarren(a)nvidia.com> --- drivers/usb/gadget/ci_udc.c | 28 ++++++++++++---------------- 1 file changed, 12 insertions(+), 16 deletions(-) diff --git a/drivers/usb/gadget/ci_udc.c b/drivers/usb/gadget/ci_udc.c index 6dc20c6c954c..5f308563e246 100644 --- a/drivers/usb/gadget/ci_udc.c +++ b/drivers/usb/gadget/ci_udc.c @@ -697,6 +697,17 @@ int usb_gadget_handle_interrupts(void) return value; } +void udc_disconnect(void) +{ + struct ci_udc *udc = (struct ci_udc *)controller.ctrl->hcor; + /* disable pullup */ + stop_activity(); + writel(USBCMD_FS2, &udc->usbcmd); + udelay(800); + if (controller.driver) + controller.driver->disconnect(&controller.gadget); +} + static int ci_pullup(struct usb_gadget *gadget, int is_on) { struct ci_udc *udc = (struct ci_udc *)controller.ctrl->hcor; @@ -715,27 +726,12 @@ static int ci_pullup(struct usb_gadget *gadget, int is_on) /* Turn on the USB connection by enabling the pullup resistor */ writel(USBCMD_ITC(MICRO_8FRAME) | USBCMD_RUN, &udc->usbcmd); } else { - stop_activity(); - writel(USBCMD_FS2, &udc->usbcmd); - udelay(800); - if (controller.driver) - controller.driver->disconnect(gadget); + udc_disconnect(); } return 0; } -void udc_disconnect(void) -{ - struct ci_udc *udc = (struct ci_udc *)controller.ctrl->hcor; - /* disable pullup */ - stop_activity(); - writel(USBCMD_FS2, &udc->usbcmd); - udelay(800); - if (controller.driver) - controller.driver->disconnect(&controller.gadget); -} - static int ci_udc_probe(void) { struct ept_queue_head *head; -- 1.8.1.5

2 4