Message ID | 20231113051929.1952351-1-mpe@ellerman.id.au (mailing list archive) |
---|---|
State | Accepted |
Commit | c8a1634145c23a5a979a7166a12b99871812a6ab |
Headers | show |
Series | powerpc/32: Drop unused grackle_set_stg() | expand |
Context | Check | Description |
---|---|---|
snowpatch_ozlabs/github-powerpc_ppctests | success | Successfully ran 8 jobs. |
snowpatch_ozlabs/github-powerpc_selftests | success | Successfully ran 8 jobs. |
snowpatch_ozlabs/github-powerpc_sparse | success | Successfully ran 4 jobs. |
snowpatch_ozlabs/github-powerpc_clang | success | Successfully ran 6 jobs. |
snowpatch_ozlabs/github-powerpc_kernel_qemu | success | Successfully ran 23 jobs. |
On Mon, 13 Nov 2023 16:19:29 +1100, Michael Ellerman wrote: > The call to grackle_set_stg() ("Store Gathering") has always been inside > an #ifdef 0, since the code was first merged in v2.3.43pre7. > > Apparently it was suspected of causing problems on some hardware so was > disabled. No one has ever proved otherwise so drop the code as unused > for now. > > [...] Applied to powerpc/next. [1/1] powerpc/32: Drop unused grackle_set_stg() https://git.kernel.org/powerpc/c/c8a1634145c23a5a979a7166a12b99871812a6ab cheers
diff --git a/arch/powerpc/sysdev/grackle.c b/arch/powerpc/sysdev/grackle.c index fd2f94a884f0..7dce8278b71e 100644 --- a/arch/powerpc/sysdev/grackle.c +++ b/arch/powerpc/sysdev/grackle.c @@ -18,24 +18,8 @@ #define GRACKLE_CFA(b, d, o) (0x80 | ((b) << 8) | ((d) << 16) \ | (((o) & ~3) << 24)) -#define GRACKLE_PICR1_STG 0x00000040 #define GRACKLE_PICR1_LOOPSNOOP 0x00000010 -/* N.B. this is called before bridges is initialized, so we can't - use grackle_pcibios_{read,write}_config_dword. */ -static inline void grackle_set_stg(struct pci_controller* bp, int enable) -{ - unsigned int val; - - out_be32(bp->cfg_addr, GRACKLE_CFA(0, 0, 0xa8)); - val = in_le32(bp->cfg_data); - val = enable? (val | GRACKLE_PICR1_STG) : - (val & ~GRACKLE_PICR1_STG); - out_be32(bp->cfg_addr, GRACKLE_CFA(0, 0, 0xa8)); - out_le32(bp->cfg_data, val); - (void)in_le32(bp->cfg_data); -} - static inline void grackle_set_loop_snoop(struct pci_controller *bp, int enable) { unsigned int val; @@ -56,7 +40,4 @@ void __init setup_grackle(struct pci_controller *hose) pci_add_flags(PCI_REASSIGN_ALL_BUS); if (of_machine_is_compatible("AAPL,PowerBook1998")) grackle_set_loop_snoop(hose, 1); -#if 0 /* Disabled for now, HW problems ??? */ - grackle_set_stg(hose, 1); -#endif }
The call to grackle_set_stg() ("Store Gathering") has always been inside an #ifdef 0, since the code was first merged in v2.3.43pre7. Apparently it was suspected of causing problems on some hardware so was disabled. No one has ever proved otherwise so drop the code as unused for now. Reported-by: kernel test robot <lkp@intel.com> Reported-by: Bjorn Helgaas <helgaas@kernel.org> Closes: https://lore.kernel.org/all/20231031145600.GA9161@bhelgaas/ Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> --- arch/powerpc/sysdev/grackle.c | 19 ------------------- 1 file changed, 19 deletions(-) Some additional detail too verbose/boring for the change log. Initial commit introducing the code, under #ifdef 0: https://git.kernel.org/pub/scm/linux/kernel/git/mpe/linux-fullhistory.git/tree/arch/ppc/kernel/pmac_pci.c?id=2fac5cbf1dd9a7720aa0130575a0b36ba4f04ede#n623 Discussion of the hardware problems it might have caused, but with no meaningful conclusion: https://lore.kernel.org/all/20000903134303.2744@192.168.1.10/