On 1 September 2017 at 14:58, Sam Protsenko <semen.protse...@linaro.org> wrote: > On 1 September 2017 at 11:19, Marek Vasut <ma...@denx.de> wrote: >> On 08/31/2017 08:20 PM, Sam Protsenko wrote: >>> Since 842778a09104 commit, "fastboot devices" stopped to show correct >>> device serial number for TI boards, showing this line instead: >>> >>> ???????????? fastboot >>> >>> This is because serial# env variable could be set after g_dnl gadget was >>> initialized (e.g. by using env_set() in the board file). >>> >>> To fix this, before checking g_dnl_serial, let's re-check if we have >>> valid serial# value. And if so, let's set it as g_dnl_serial value. >>> >>> Signed-off-by: Sam Protsenko <semen.protse...@linaro.org> >> >> Can't we use U_BOOT_ENV_CALLBACK() here instead and set the serial >> number when the environment variable is set ? >> > > Good point. I've tested this approach and it works fine. Will resend > v2 shortly. Thanks for the hint! >
Sent v2, using U_BOOT_ENV_CALLBACK: [1]. Guys, please review. [1] https://lists.denx.de/pipermail/u-boot/2017-September/304460.html >>> --- >>> drivers/usb/gadget/g_dnl.c | 9 +++++++++ >>> 1 file changed, 9 insertions(+) >>> >>> diff --git a/drivers/usb/gadget/g_dnl.c b/drivers/usb/gadget/g_dnl.c >>> index 0491a0eea9..e4d0289757 100644 >>> --- a/drivers/usb/gadget/g_dnl.c >>> +++ b/drivers/usb/gadget/g_dnl.c >>> @@ -226,6 +226,15 @@ static int g_dnl_bind(struct usb_composite_dev *cdev) >>> >>> g_dnl_bind_fixup(&device_desc, cdev->driver->name); >>> >>> + /* First try to obtain serial number from serial# variable */ >>> + if (strlen(g_dnl_serial) == 0) { >>> + const char *s; >>> + >>> + s = env_get("serial#"); >>> + if (s) >>> + g_dnl_set_serialnumber((char *)s); >>> + } >>> + >>> if (strlen(g_dnl_serial)) { >>> id = usb_string_id(cdev); >>> if (id < 0) >>> >> >> >> -- >> Best regards, >> Marek Vasut _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de https://lists.denx.de/listinfo/u-boot
