Message ID | 20220906074509.928865-1-ajones@ventanamicro.com (mailing list archive) |
---|---|
Headers | show |
Series | Zicbom block size fixes and cleanups | expand |
On Tue, 06 Sep 2022 00:45:07 PDT (-0700), ajones@ventanamicro.com wrote: > I collected a couple patches that were necessary for another series. > Since some changes were required to the second patch I'm posting in > case it simplifies things. > > v3: > - Applied Anup's move change first to ensure the second patch can > compile with Heiko's change. > > v2: > - First attempt at sending a fixed-up version of the second patch, > which mistakenly included Heiko's change without Anup's change. > > v1: > - The original patches on the list > 1. [PATCH v2 2/4] RISC-V: Move riscv_init_cbom_blocksize() to cacheflush.c > 20220829125226.511564-5-apatel@ventanamicro.com > 2. [PATCH v2] RISC-V: Clean up the Zicbom block size probing > 20220812154010.18280-1-palmer@rivosinc.com > > Anup Patel (1): > RISC-V: Move riscv_init_cbom_blocksize() to cacheflush.c > > Palmer Dabbelt (1): > RISC-V: Clean up the Zicbom block size probing Sorry I dropped the ball on this one, but having the fix depend on a refactoring isn't the right way to go as those refactorings aren't supposed to end up on fixes. I merged it without the move into palmer/riscv-zicbom_fix, but the jet lag is starting to kick in so I'm going to not merge it onto fixes for this week. > > arch/riscv/errata/thead/errata.c | 1 + > arch/riscv/include/asm/cacheflush.h | 2 ++ > arch/riscv/kernel/setup.c | 2 +- > arch/riscv/mm/cacheflush.c | 40 +++++++++++++++++++++++++++++ > arch/riscv/mm/dma-noncoherent.c | 40 ++--------------------------- > 5 files changed, 46 insertions(+), 39 deletions(-)
On 09/09/2022 00:14, Palmer Dabbelt wrote: > On Tue, 06 Sep 2022 00:45:07 PDT (-0700), ajones@ventanamicro.com > wrote: >> I collected a couple patches that were necessary for another >> series. Since some changes were required to the second patch I'm >> posting in case it simplifies things. >> >> v3: - Applied Anup's move change first to ensure the second patch >> can compile with Heiko's change. >> >> v2: - First attempt at sending a fixed-up version of the second >> patch, which mistakenly included Heiko's change without Anup's >> change. >> >> v1: - The original patches on the list 1. [PATCH v2 2/4] RISC-V: >> Move riscv_init_cbom_blocksize() to cacheflush.c >> 20220829125226.511564-5-apatel@ventanamicro.com 2. [PATCH v2] >> RISC-V: Clean up the Zicbom block size probing >> 20220812154010.18280-1-palmer@rivosinc.com >> >> Anup Patel (1): RISC-V: Move riscv_init_cbom_blocksize() to >> cacheflush.c >> >> Palmer Dabbelt (1): RISC-V: Clean up the Zicbom block size probing > > Sorry I dropped the ball on this one, but having the fix depend on a > refactoring isn't the right way to go as those refactorings aren't > supposed to end up on fixes. I merged it without the move into > palmer/riscv-zicbom_fix, but the jet lag is starting to kick in so > I'm going to not merge it onto fixes for this week. Good call! /stuff/linux/arch/riscv/mm/dma-noncoherent.c:15:14: error: redefinition of 'riscv_cbom_block_size' with a different type: 'unsigned int' vs 'int' unsigned int riscv_cbom_block_size; ^ /stuff/linux/arch/riscv/include/asm/cacheflush.h:46:12: note: previous declaration is here extern int riscv_cbom_block_size; ^ 1 error generated. clang-15 w/ zicbom CONFIG_RISCV_ISA_ZICBOM=y Hope that's helpful.. Conor. > >> >> arch/riscv/errata/thead/errata.c | 1 + >> arch/riscv/include/asm/cacheflush.h | 2 ++ >> arch/riscv/kernel/setup.c | 2 +- >> arch/riscv/mm/cacheflush.c | 40 >> +++++++++++++++++++++++++++++ arch/riscv/mm/dma-noncoherent.c | >> 40 ++--------------------------- 5 files changed, 46 insertions(+), >> 39 deletions(-) > > _______________________________________________ linux-riscv mailing > list linux-riscv@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-riscv