https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64953
Bug ID: 64953
Summary: Compiling sourcecode for STM32F103 causes USB errors
with some optimization settings
Product: gcc
Version: 4.9.2
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: manuel.reimer at gmx dot de
I've tried to compile this sourcecode for an STM32F103 controller:
https://github.com/tormodvolden/maple-bootloader/tree/unify_platforms
I've modified the Makefile to use "-Os" to reduce size used in flash.
This worked well, so far, with an gcc 4.8 compiler. But if I do this with a
4.9.2 compiler, then the USB handshake fails with ugly errors on dmesg:
[ 2068.103728] usb 1-1: new full-speed USB device number 2 using xhci_hcd
[ 2068.217014] usb 1-1: device descriptor read/64, error -71
[ 2068.323573] xhci_hcd 0000:02:00.0: Setup ERROR: setup context command for
slot 1.
[ 2068.327349] usb 1-1: hub failed to enable device, error -22
[ 2068.436815] usb 1-1: new full-speed USB device number 3 using xhci_hcd
[ 2068.550102] usb 1-1: device descriptor read/64, error 18
[ 2068.656663] xhci_hcd 0000:02:00.0: Setup ERROR: setup context command for
slot 1.
[ 2068.660448] usb 1-1: hub failed to enable device, error -22
[ 2068.773235] usb 1-1: new full-speed USB device number 4 using xhci_hcd
[ 2068.773866] xhci_hcd 0000:02:00.0: ERROR: unexpected setup address command
completion code 0x6.
[ 2068.980792] xhci_hcd 0000:02:00.0: ERROR: unexpected setup address command
completion code 0x6.
[ 2069.186265] usb 1-1: device not accepting address 4, error -22
[ 2069.296195] usb 1-1: new full-speed USB device number 5 using xhci_hcd
[ 2069.296483] xhci_hcd 0000:02:00.0: ERROR: unexpected setup address command
completion code 0x6.
[ 2069.503332] xhci_hcd 0000:02:00.0: ERROR: unexpected setup address command
completion code 0x6.
[ 2069.709220] usb 1-1: device not accepting address 5, error -22
[ 2069.713114] usb usb1-port1: unable to enumerate USB device
After some trying I've found out that the key in this case is the optimization
setting. If I use "-O3", then the resulting binary works again if I use the
4.9.2 compiler.
So "something" is wrong with the optimization.
Can someone give me a hint about what changed between 4.8 and 4.9.2? Is this a
bug in the gcc compiler or do I just need some new "compiler switch" to disable
"some new feature" and get a valid binary, again?
