libusbx v1.0.9-rc4 is now available

[Sean McBride]

On Fri, 30 Mar 2012 13:49:04 +0100, Pete Batard said:

>---
> libusb/os/darwin_usb.c |    2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
>diff --git a/libusb/os/darwin_usb.c b/libusb/os/darwin_usb.c
>index e112cf5..c51557a 100644
>--- a/libusb/os/darwin_usb.c
>+++ b/libusb/os/darwin_usb.c
>@@ -1547,7 +1547,7 @@ static void darwin_async_io_callback (void
>*refcon, IOReturn result, void *arg0)
> 
>   usbi_info (ITRANSFER_CTX (itransfer), "an async io operation has
>completed");
> 
>-  size = (UInt32) arg0;
>+  size = (UInt32) (intptr_t) arg0;
> 
>   /* send a completion message to the device's file descriptor */
>   message = MESSAGE_ASYNC_IO_COMPLETE;
>-- 

Please do not merge this change!

You're not fixing anything, you're just tricking the compiler into not warning you about a real problem.  You're still throwing away half the bits of arg0 and now you've also discarded the automated TODO the compiler was giving us.

This is bug #117:
<http://www.libusb.org/ticket/117>

That warning was deliberately introduced:
<http://www.libusb.org/attachment/ticket/117/0089-conversative-partial-fix-for-ticket-117.patch>

Unfortunately the big fat warning comment I added in my patch was apparently removed.  Blah!

IMNSHO:
 - the warning should stay
 - a TODO comment should be added referencing bug #117
 - an assert should check that size is < UINT32_MAX

(I'd submit a patch, but I consider libusb a waste of my time at this point.  I hope libusbx will change that, but am not yet convinced.)

-- 
____________________________________________________________
Sean McBride, B. Eng                 sean at rogue-research.com
Rogue Research                        www.rogue-research.com 
Mac Software Developer              Montréal, Québec, Canada