imx_usb for bulk devices broken?

[agners]

Hello,

I am working on cleaning up the code a bit. I came across the following strange statement:

if (p_id->mode == MODE_HID) if (type == FT_APP) {

It leads to the fact that jump command is not executed for bulk devices (MX51/MX53).

I think this creept in accidentially since in that stament there is another if which checks for HID, and since the jump command seemed to be sent unconditionally before:

166097f

Anybody uses imx_usb_loader with such a device recently?

[tkisky]

On 2/9/2018 7:54 AM, Stefan Agner wrote: Hello, I am working on cleaning up the code a bit. I came across the following strange statement: |if (p_id->mode == MODE_HID) if (type == FT_APP) { | It leads to the fact that jump command is not executed for bulk devices (MX51/MX53). I think this creept in accidentially since in that stament there is another if which checks for HID, and since the jump command seemed to be sent unconditionally before: 166097f <166097f> Anybody uses imx_usb_loader with such a device recently? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#70>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABoNv9FLEgbj2Q6nJ7cHIDuh0Ndut0pAks5tTGosgaJpZM4SAID0>.

I've used on an imx51 on 138c0b2 That inside if, that is always true now, used to be always false - if (0) if (dev->mode == MODE_HID) { + if (dev->mode == MODE_HID) { commit 6deb910 Author: Jeremie Corbier <[email protected]> Date: Tue Apr 19 10:58:06 2016 +0200 Use the DCD_WRITE SDP command Signed-off-by: Jeremie Corbier <[email protected]>

…

__________________________ We need to get better at 1 patch doing only 1 thing. But, remember this comment //Any command will initiate jump for mx51, jump address is ignored by mx51 And jumps are no longer very relevant. Troy

[agners]

Hm, what is exaclty the conclusion here? It works even without any command for mx51 (except download before, I suppose?).

The (reenabled) report 4 is helpful in error cases and correct IMHO...