Re: [AArch64] big.LITTLE option rewriting bug

2014-01-07 Thread Marcus Shawcroft
On 07/01/14 17:02, James Greenhalgh wrote: Ugh. Now we have two ASM_SPECs which try to handle -mcpu as input. One of them just returns the input, the other does the cpu rewriting we actually want, so we can end up with: gcc -mcpu=cortex-a57.cortex-a53 Getting passed through to the assembl

Re: [AArch64] big.LITTLE option rewriting bug

2014-01-07 Thread Richard Earnshaw
On 07/01/14 17:02, James Greenhalgh wrote: > > Ugh. > > Now we have two ASM_SPECs which try to handle -mcpu as input. > > One of them just returns the input, the other does the cpu rewriting > we actually want, so we can end up with: > > gcc -mcpu=cortex-a57.cortex-a53 > > Getting passed thr