Re: [RFC PATCH 1/7] powerpc: use 16-bit immediate for STACK_FRAME_REGS_MARKER

2022-09-23 Thread Christophe Leroy
Le 23/09/2022 à 09:32, Michael Ellerman a écrit : > Christophe Leroy writes: >> Le 19/09/2022 à 16:01, Nicholas Piggin a écrit : >>> Using a 16-bit constant for this marker allows it to be loaded with >>> a single 'li' instruction. On 64-bit this avoids a TOC entry and a >>> TOC load that

Re: [RFC PATCH 1/7] powerpc: use 16-bit immediate for STACK_FRAME_REGS_MARKER

2022-09-23 Thread Michael Ellerman
Christophe Leroy writes: > Le 19/09/2022 à 16:01, Nicholas Piggin a écrit : >> Using a 16-bit constant for this marker allows it to be loaded with >> a single 'li' instruction. On 64-bit this avoids a TOC entry and a >> TOC load that depends on the r2 value that has just been loaded from >> the

Re: [RFC PATCH 1/7] powerpc: use 16-bit immediate for STACK_FRAME_REGS_MARKER

2022-09-21 Thread Christophe Leroy
Le 19/09/2022 à 16:01, Nicholas Piggin a écrit : > Using a 16-bit constant for this marker allows it to be loaded with > a single 'li' instruction. On 64-bit this avoids a TOC entry and a > TOC load that depends on the r2 value that has just been loaded from > the PACA. > > XXX: this probably

[RFC PATCH 1/7] powerpc: use 16-bit immediate for STACK_FRAME_REGS_MARKER

2022-09-19 Thread Nicholas Piggin
Using a 16-bit constant for this marker allows it to be loaded with a single 'li' instruction. On 64-bit this avoids a TOC entry and a TOC load that depends on the r2 value that has just been loaded from the PACA. XXX: this probably should be 64-bit change and use 2 instruction sequence that