Re: [U-Boot] [RFC] powerpc/lib: unsafe register handling in wait_ticks

2013-01-25 Thread Joakim Tjernlund
Wolfgang Denk w...@denx.de wrote on 2013/01/24 20:27:26: Dear Joakim Tjernlund, In message OF52C94A3D.C3BD2E6F-ONC1257AFD.005BAFE0-C1257AFD. 00673...@transmode.se you wrote: This adds some extra churn to the code that my patch didn't do. On the other hand your patch makes the

[U-Boot] [RFC] powerpc/lib: unsafe register handling in wait_ticks

2013-01-24 Thread Mats Kärrman
Hi, If watchdog is enabled, the arch/powerpc/lib/ticks.S::wait_ticks() function calls the function specified by the WATCHDOG_RESET macro. The wait_ticks function depends on the registers r0, r6 and r7 being preserved however that is not guaranteed if the reset function is a C function. The

Re: [U-Boot] [RFC] powerpc/lib: unsafe register handling in wait_ticks

2013-01-24 Thread Joakim Tjernlund
Hi, Hi Mats I would appreciate if you CC me directly on stuff I have been involved in. I don't read every mail on u-boot list(to many of them). It was just plain luck I saw this one. If watchdog is enabled, the arch/powerpc/lib/ticks.S::wait_ticks() function calls the function

Re: [U-Boot] [RFC] powerpc/lib: unsafe register handling in wait_ticks

2013-01-24 Thread Wolfgang Denk
Dear Joakim Tjernlund, In message of52c94a3d.c3bd2e6f-onc1257afd.005bafe0-c1257afd.00673...@transmode.se you wrote: This adds some extra churn to the code that my patch didn't do. On the other hand your patch makes the function look more like how gcc would have done so I am fine with that.