diff options
Diffstat (limited to 'include')
| -rw-r--r-- | include/asm-ia64/io.h | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/include/asm-ia64/io.h b/include/asm-ia64/io.h index cc2f3d6803c3..2ff2d00ef3ed 100644 --- a/include/asm-ia64/io.h +++ b/include/asm-ia64/io.h @@ -414,6 +414,16 @@ extern void __ia64_memset_c_io (unsigned long, unsigned long, long); # endif /* __KERNEL__ */ /* + * Enabling BIO_VMERGE_BOUNDARY forces us to turn off I/O MMU bypassing. It is said that + * BIO-level virtual merging can give up to 4% performance boost (not verified for ia64). + * On the other hand, we know that I/O MMU bypassing gives ~8% performance improvement on + * SPECweb-like workloads on zx1-based machines. Thus, for now we favor I/O MMU bypassing + * over BIO-level virtual merging. + */ +#if 1 +#define BIO_VMERGE_BOUNDARY 0 +#else +/* * It makes no sense at all to have this BIO_VMERGE_BOUNDARY macro here. Should be * replaced by dma_merge_mask() or something of that sort. Note: the only way * BIO_VMERGE_BOUNDARY is used is to mask off bits. Effectively, our definition gets @@ -425,5 +435,6 @@ extern void __ia64_memset_c_io (unsigned long, unsigned long, long); */ extern unsigned long ia64_max_iommu_merge_mask; #define BIO_VMERGE_BOUNDARY (ia64_max_iommu_merge_mask + 1) +#endif #endif /* _ASM_IA64_IO_H */ |