- 23 Apr, 2013 8 commits
-
-
Shawn Guo authored
Commit edc7cb2e (usb: phy: make it a menuconfig) makes USB_MXS_PHY be a sub-item of menuconfig symbol USB_PHY. This change gets the selection of CONFIG_USB_MXS_PHY in mxs_defconfig lost. Hence the boot stops at the point below. [ 1.600867] ci_hdrc ci_hdrc.0: doesn't support gadget [ 1.606282] ci_hdrc ci_hdrc.0: EHCI Host Controller [ 1.613522] ci_hdrc ci_hdrc.0: new USB bus registered, assigned bus number 1 Add CONFIG_USB_PHY to have the CONFIG_USB_MXS_PHY selection back to work. Signed-off-by:
Shawn Guo <shawn.guo@linaro.org> Signed-off-by:
Felipe Balbi <balbi@ti.com>
-
Shawn Guo authored
Commit edc7cb2e (usb: phy: make it a menuconfig) makes USB_MXS_PHY be a sub-item of menuconfig symbol USB_PHY. This change gets the selection of CONFIG_USB_MXS_PHY in imx_v6_v7_defconfig lost. Hence the boot stops at the point below. ci_hdrc ci_hdrc.0: doesn't support gadget ci_hdrc ci_hdrc.0: EHCI Host Controller ci_hdrc ci_hdrc.0: new USB bus registered, assigned bus number 1 Add CONFIG_USB_PHY to have the CONFIG_USB_MXS_PHY selection back to work. Signed-off-by:
-
Denis Efremov authored
The symbol usb_bind_phy is exported and annotated __init. It looks like section mismatch. Fix by removing the __init annotation of usb_bind_phy. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by:
Denis Efremov <yefremov.denis@gmail.com> Signed-off-by:
-
Andrzej Pietrasiewicz authored
If function instances are not put on gadget's unbind, their implementation module's refcount is nonzero and it is impossible to unload them. Signed-off-by:
Andrzej Pietrasiewicz <andrzej.p@samsung.com> Signed-off-by:
Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by:
-
Andrzej Pietrasiewicz authored
acm was the first function to be converted and it seems that its code served as a base for converting f_sourcesink to the new function interface. source_sink has nothing to do with acm, though. Signed-off-by:
-
Wei Yongjun authored
Fix to return a negative error code from the error handling case instead of 0, as returned elsewhere in this function. Introduced by commit 29a664 (usb: gadget: cdc2: use function framework for ACM) Signed-off-by:
Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by:
-
Wei Yongjun authored
Fix to return a negative error code from the error handling case instead of 0, as returned elsewhere in this function. Introduced by commit 59835a (usb: gadget: multi: use function framework for ACM.) Signed-off-by:
-
Wei Yongjun authored
Fix to return a negative error code from the error handling case instead of 0, as returned elsewhere in this function. Signed-off-by:
-
- 03 Apr, 2013 30 commits
-
-
Felipe Balbi authored
recent changes to gadget framework have exposed a flaw in nokia.c where it uses f_ecm.c but it wasn't including it like other gadget drivers. This is a temporary patch until all other function drivers are converted to new API where each function driver becomes a module and binding is done through configfs. Solves a linking error when building g_nokia.ko Signed-off-by: -
Fabio Baltieri authored
Since regulator_enable() is going to be marked as __must_check in the next merge window, always check regulator_enable() return value and print a warning if it fails. Cc: Hema HK <hemahk@ti.com> Signed-off-by:
Fabio Baltieri <fabio.baltieri@linaro.org> Signed-off-by:
-
Fabio Baltieri authored
Since regulator_enable() is going to be marked as __must_check in the next merge window, always check regulator_enable() return value and print a warning if it fails. Reviewed-by:
Kalle Jokiniemi <kalle.jokiniemi@jollamobile.com> Signed-off-by:
-
Fabio Baltieri authored
Since regulator_enable() is going to be marked as __must_check in the next merge window, always check regulator_enable() return value and print a warning if it fails. Cc: Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
-
Andrzej Pietrasiewicz authored
f_obex learns about our new configfs-based interface, which will allow gadgets to be bound to controllers through userland. Signed-off-by:
-
Andrzej Pietrasiewicz authored
f_obex is now a self-contained module. We need to teach serial.c about it. Signed-off-by:
-
Andrzej Pietrasiewicz authored
Converting f_obex to the new function interface requires converting the f_obex's function code and its users. This patch converts the f_obex.c to the new function interface. The file is now compiled into a separate usb_f_obex.ko module. The old function interface is provided by means of preprocessor conditional directives. After all users are converted, the old interface can be removed. Signed-off-by:
-
Andrzej Pietrasiewicz authored
this patch implements the new configfs based interface on f_serial function driver. Signed-off-by:
-
Andrzej Pietrasiewicz authored
There are no old function interface users left, so the old interface can be removed. Signed-off-by:
-
Andrzej Pietrasiewicz authored
Signed-off-by:
-
Andrzej Pietrasiewicz authored
Converting f_serial to the new function interface requires converting the f_serial's function code and its users. This patch converts the f_serial.c to the new function interface. The file is now compiled into a separate usb_f_serial.ko module. The old function interface is provided by means of preprocessor conditional directives. After all users are converted, the old interface can be removed. Signed-off-by:
-
Andrzej Pietrasiewicz authored
f_serial.c isn't necessary in nokia.c, we can safely remove it. Signed-off-by:
-
Andrzej Pietrasiewicz authored
In order to convert to configfs the USB functions need to be converted to the new interface from Sebastian, which also requires turning them into separate modules. Some of these modules will consist of just one object file, e.g. f_ncm.o. But some of the modules will eventually consist of more than one object file, e.g. for mass storage there will be f_mass_storage.o and storage_common.o. The resulting module cannot be called f_mass_storage.ko due to cyclic dependency. This patch introduces a naming scheme for the said resulting modules: usb_f_xxxxxx.ko e.g. usb_f_mass_storage.ko, usb_f_ss_lb.ko, usb_f_rndis.ko etc. Signed-off-by:
-
Sebastian Andrzej Siewior authored
|# modprobe dummy_hcd num=2 |# modprobe libcomposite |# lsmod |Module Size Used by |libcomposite 31648 0 |dummy_hcd 19871 0 |# mkdir /sys/kernel/config/usb_gadget/oha |# cd /sys/kernel/config/usb_gadget/oha |# mkdir configs/def.1 |# mkdir configs/def.2 |# mkdir functions/acm.ttyS1 |# mkdir strings/0x1 |mkdir: cannot create directory `strings/0x1': Invalid argument |# mkdir strings/0x409 |# mkdir strings/1033 |mkdir: cannot create directory `strings/1033': File exists |# mkdir strings/1032 |# mkdir configs/def.1/strings/0x409 |# mkdir configs/def.2/strings/0x409 |#find . -ls | 975 0 drwxr-xr-x 5 root root 0 Dec 23 17:40 . | 978 0 drwxr-xr-x 4 root root 0 Dec 23 17:43 ./strings | 4100 0 drwxr-xr-x 2 root root 0 Dec 23 17:43 ./strings/1032 | 995 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/1032/serialnumber | 996 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/1032/product | 997 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/1032/manufacturer | 2002 0 drwxr-xr-x 2 root root 0 Dec 23 17:41 ./strings/0x409 | 998 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/0x409/serialnumber | 999 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/0x409/product | 1000 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./strings/0x409/manufacturer | 977 0 drwxr-xr-x 4 root root 0 Dec 23 17:41 ./configs | 4081 0 drwxr-xr-x 3 root root 0 Dec 23 17:41 ./configs/def.2 | 4082 0 drwxr-xr-x 3 root root 0 Dec 23 17:42 ./configs/def.2/strings | 2016 0 drwxr-xr-x 2 root root 0 Dec 23 17:42 ./configs/def.2/strings/0x409 | 1001 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.2/strings/0x409/configuration | 1002 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.2/bmAttributes | 1003 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.2/MaxPower | 979 0 drwxr-xr-x 3 root root 0 Dec 23 17:42 ./configs/def.1 | 980 0 drwxr-xr-x 3 root root 0 Dec 23 17:42 ./configs/def.1/strings | 5122 0 drwxr-xr-x 2 root root 0 Dec 23 17:42 ./configs/def.1/strings/0x409 | 1004 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.1/strings/0x409/configuration | 1005 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.1/bmAttributes | 1006 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./configs/def.1/MaxPower | 976 0 drwxr-xr-x 3 root root 0 Dec 23 17:41 ./functions | 981 0 drwxr-xr-x 2 root root 0 Dec 23 17:41 ./functions/acm.ttyS1 | 1007 0 -r--r--r-- 1 root root 4096 Dec 23 17:43 ./functions/acm.ttyS1/port_num | 1008 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./UDC | 1009 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bcdUSB | 1010 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bcdDevice | 1011 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./idProduct | 1012 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./idVendor | 1013 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bMaxPacketSize0 | 1014 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bDeviceProtocol | 1015 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bDeviceSubClass | 1016 0 -rw-r--r-- 1 root root 4096 Dec 23 17:43 ./bDeviceClass |# cat functions/acm.ttyS1/port_num |0 |# ls -lah /dev/ttyGS* |crw-rw---T 1 root dialout 252, 0 Dec 23 17:41 /dev/ttyGS0 | |# echo 0x1234 > idProduct |# echo 0xabcd > idVendor |# echo 1122 > strings/0x409/serialnumber |# echo "The manufacturer" > strings/0x409/manufacturer |# echo 1 > strings/1032/manufacturer |# echo 1sa > strings/1032/product |# echo tada > strings/1032/serialnumber |echo "Primary configuration" > configs/def.1/strings/0x409/configuration |# echo "Secondary configuration" > configs/def.2/strings/0x409/configuration |# ln -s functions/acm.ttyS1 configs/def.1/ |# ln -s functions/acm.ttyS1 configs/def.2/ |find configs/def.1/ -ls | 979 0 drwxr-xr-x 3 root root 0 Dec 23 17:49 configs/def.1/ | 6264 0 lrwxrwxrwx 1 root root 0 Dec 23 17:48 configs/def.1/acm.ttyS1 -> ../../../../usb_gadget/oha/functions/acm.ttyS1 | 980 0 drwxr-xr-x 3 root root 0 Dec 23 17:42 configs/def.1/strings | 5122 0 drwxr-xr-x 2 root root 0 Dec 23 17:49 configs/def.1/strings/0x409 | 6284 0 -rw-r--r-- 1 root root 4096 Dec 23 17:47 configs/def.1/strings/0x409/configuration | 6285 0 -rw-r--r-- 1 root root 4096 Dec 23 17:49 configs/def.1/bmAttributes | 6286 0 -rw-r--r-- 1 root root 4096 Dec 23 17:49 configs/def.1/MaxPower | |echo 120 > configs/def.1/MaxPower | |# ls -lh /sys/class/udc/ |total 0 |lrwxrwxrwx 1 root root 0 Dec 23 17:50 dummy_udc.0 -> ../../devices/platform/dummy_udc.0/udc/dummy_udc.0 |lrwxrwxrwx 1 root root 0 Dec 23 17:50 dummy_udc.1 -> ../../devices/platform/dummy_udc.1/udc/dummy_udc.1 |# echo dummy_udc.0 > UDC |# lsusb |Bus 001 Device 002: ID abcd:1234 Unknown | |lsusb -d abcd:1234 -v |Device Descriptor: … | idVendor 0xabcd Unknown | idProduct 0x1234 | bcdDevice 3.06 | iManufacturer 1 The manufacturer | iProduct 2 | iSerial 3 1122 | bNumConfigurations 2 … |echo "" > UDC v5…v6 - wired up strings with usb_gstrings_attach() - add UDC attribe. Write "udc-name" will bind the gadget. Write an empty string (it should contain \n since 0 bytes write get optimzed away) will unbind the UDC from the gadget. The name of available UDCs can be obtained from /sys/class/udc/ v4…v5 - string rework. This will add a strings folder incl. language code like strings/409/manufacturer as suggested by Alan. - rebased ontop reworked functions.c which has usb_function_instance which is used prior after "mkdir acm.instance" and can be directly used for configuration via configfs. v3…v4 - moved functions from the root folde down to the gadget as suggested by Michał - configs have now their own configs folder as suggested by Michał. The folder is still name.bConfigurationValue where name becomes the sConfiguration. Is this usefull should we just stilc configs/bConfigurationValue/ ? - added configfs support to the ACM function. The port_num attribute is exported by f_acm. An argument has been added to the USB alloc function to distinguish between "old" (use facm_configure() to configure and configfs interface (expose a config_node). The port_num is currently a dumb counter. It will require some function re-work to make it work. scheduled for v5: - sym linking function into config. v2…v3 - replaced one ifndef by ifdef as suggested by Micahał - strstr()/strchr() function_make as suggested by Micahł - replace [iSerialNumber|iProduct|iManufacturer] with [sSerialNumber|sProduct|sManufacturer] as suggested by Alan - added creation of config descriptors v1…v2 - moved gadgets from configfs' root directory into /udcs/ within our "usb_gadget" folder. Requested by Andrzej & Michał - use a dot as a delimiter between function's name and its instance's name as suggested by Michał - renamed all config_item_type, configfs_group_operations, make_group, drop_item as suggested by suggested by Andrzej to remain consisten within this file and within other configfs users - Since configfs.c and functions.c are now part of the udc-core module, the module itself is now called udc. Also added a tiny ifdef around init code becuase udc-core is subsys init and this is too early for configfs in the built-in case. In the module case, we can only have one init function. Signed-off-by:Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by:
-
Sebastian Andrzej Siewior authored
It possible to allocate the tty port number within the "instance" structure of the function and there is no need to expose this information within the gadget and therefore it is removed here. This patch converts only f_acm and all its users. The other gadgets will follow once the function is converted to the function interface. Signed-off-by:
-
Sebastian Andrzej Siewior authored
the u_ether.c file has a global variable named the_dev which keeps a pointer to the network device after it has been created via gether_setup_name(). It is only used internally by u_ether. This patches moves the variable to its users and passes it via the port.ioport where it is saved later anyway. Signed-off-by:
-
Sebastian Andrzej Siewior authored
This patch converts the acm_ms gadget to make use of the function framework to request the ACM function. The "old" include interface for acm is now removed since nokia was the last user of it (for ACM). Signed-off-by:
-
Sakethram Bommisetti authored
Modify ab8500_usb_set_peripheral() and ab8500_usb_set_host() code to schedule phy_dis_work only when necessary in order to prevent regulator count mismatch during reboot/shutdown. Signed-off-by:
Sakethram Bommisetti <sakethram.bommisetti@stericsson.com> Acked-by:
Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
-
Sakethram Bommisetti authored
ab8500_usb_delayed_work was implemented as a workaroud for the internal only and now unsupported v1.0 version of AB850. This patch removes the delayed work and just leave a link status update call at probe time. Signed-off-by:
-
Patrice Chotard authored
Amend the ab8500-usb driver to optionally take a pin control handle and set the state of the pins to "default" on ab8500_usb_phy_enable and to "sleep" on ab8500_usb_phy_disable. The pinctrl handle is released on ab8500_usb_phy_disable because USB pins are shared with ab8505_micro_usb_iddet driver. Signed-off-by:
Patrice Chotard <patrice.chotard@st.com> Acked-by:
-
Fabio Baltieri authored
Fix last ab->phy.notifier call to use vbus_draw as notifier argument, as that's used in ab8500_charger to control charging current. Also drop a related TODO comment, and the additional ux500_musb_set_vbus(musb, 0), as with this patch it was causing an erratic behaviour of gadget ep0 state machine. Acked-by:
-
Mian Yousaf Kaukab authored
To prevent clock and regulator frameworks from complaining, only disable the host or peripheral phy if they were enabled. Reported-by:
Mian Yousaf Kaukab <mian.yousaf.kaukab@stericsson.com> Acked-by:
-
Fabio Baltieri authored
Add ab8500_usb_regulator_{enable,disable} functions to control USB phy regulators on corresponding ab8500_usb_phy_{enable,disable} events. This contains some workaround and optimization for specific AB8500 versions. Signed-off-by:Praveena Nadahally <praveen.nadahally@stericsson.com> Acked-by:
-
Fabio Baltieri authored
Split ab8500_usb_phy_ctrl into separate enable/disable functions to make the code more linear and readable. Acked-by:
-
Fabio Baltieri authored
Add initial regulator support to ab8500-usb by introducing necessary devm_regulator_get(). Acked-by:
-
Sakethram Bommisetti authored
AB8500 v2.0 has eye diagram issues when drawing more than 100mA from VBUS. Force charging current to 100mA in case of standard host. Signed-off-by:
-
Sakethram Bommisetti authored
Set phy tuning values proposed by the hardware teams for AB8500 and AB8505 to improve USB eye diagram performances. Signed-off-by:
-
Fabio Baltieri authored
Convert local data allocation to devm_kzalloc and drop unnecessary fail path code. Acked-by:
-
Fabio Baltieri authored
Drop references to deprecated U5500 platform in driver comments. Acked-by:
-
Sachin Kamat authored
Use the newly introduced devm_ioremap_resource() instead of devm_request_and_ioremap() which provides more consistent error handling. devm_ioremap_resource() provides its own error messages; so all explicit error messages can be removed from the failure code paths. Signed-off-by:
Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by:
-
- 02 Apr, 2013 2 commits
-
-
Felipe Balbi authored
Remove all section annotations to fix the following section mismatches: >> WARNING: drivers/usb/gadget/pxa27x_udc.o(.text+0x597c): Section mismatch in reference from the function .pxa_udc_probe() to the function .init.text:.udc_init_data.constprop.11() The function .pxa_udc_probe() references the function __init .udc_init_data.constprop.11(). This is often because .pxa_udc_probe lacks a __init annotation or the annotation of .udc_init_data.constprop.11 is wrong. -- >> WARNING: drivers/usb/gadget/pxa27x_udc.o(.text+0x5a04): Section mismatch in reference from the function .pxa_udc_probe() to the function .init.text:.pxa_ep_setup() The function .pxa_udc_probe() references the function __init .pxa_ep_setup(). This is often because .pxa_udc_probe lacks a __init annotation or the annotation of .pxa_ep_setup is wrong. Reported-by:
Fengguang Wu <fengguang.wu@intel.com> Signed-off-by:
-
Felipe Balbi authored
udc-core would leak 'udc' memory in some error cases. Fix it by kfree()ing udc on error path. Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
