Any time the kernel sets up more than one virtual mapping for the same physical page, cache line aliasing may occur. The kernel is careful to avoid aliasing, so it usually occurs only in one particular instance: when the user mmaps a file. Here, the kernel has one virtual address for pages of the file in the page cache, and the user may have one or more different virtual addresses. This is possible because nothing prevents the user from mmaping the file at multiple locations.
When a file is mmaped, the kernel adds the mapping to one of the inode's lists: i_mmap, for maps that cannot change the underlying data, or i_mmap_shared, for maps that can change the file's data. The API for bringing the cache aliases of a page into sync is:
void flush_dcache_page(struct page *page);
This API must be called every time data on the page is altered by the kernel, and it should be called before reading data from the page if page->mapping->i_mmap_shared is not empty. In architecture-specific code, flush_dcache_page loops over the i_mmap_shared list and flushes the cache data. It then loops over the i_mmap list and invalidates it, thus bringing all the aliases into sync.
In their quest for efficiency, processors often have separate caches for the instructions they execute and the data on which they operate. Often, these caches are separate mechanisms, and a data write may not be seen by the instruction cache. This causes problems if you are trying to execute instructions you just wrote into memory, for example, during module loading or when using a trampoline. You must use the following API:
void flush_icache_range(unsigned long start, unsigned long end);
to ensure that the instructions are seen by the instruction cache prior to execution. start and end are the starting and ending addresses, respectively, of the block of memory you modified to contain your instructions.
These flush all the caches in the system and only the lines in the cache belonging to the particular process address space mm. Both of these are extremely expensive operations and should be used only when absolutely necessary.
When more than one CPU is in the system, a level of caching usually exists that is unique to each CPU. Depending on the architecture, it may be the responsibility of the kernel to ensure that changes in the cache of one CPU become visible to the other CPUs. Fortunately, most CPUs handle this type of coherency problem in hardware. Even if they don't, as long as you follow the APIs listed in this article, you can maintain coherency across all the CPUs.
I hope I've given you a brief overview of how caches work and how the kernel manages them. The contents of this article should be sufficient for you to understand caching in most kernel programming situations you're likely to encounter. Be aware, however, that if you get deeply into the guts of kernel cache management (particularly in the architecture-specific code), you likely will come across concepts and APIs not discussed here.
James Bottomley is the software architect for SteelEye. He maintains the SCSI subsystem, the Linux Voyager port and the 53c700 driver. He also has made contributions to PA-RISC Linux development in the area of DMA/device model abstraction.
- March 2015 Issue of Linux Journal: System Administration
- High-Availability Storage with HA-LVM
- DNSMasq, the Pint-Sized Super Dæmon!
- Localhost DNS Cache
- Real-Time Rogue Wireless Access Point Detection with the Raspberry Pi
- Days Between Dates: the Counting
- The Usability of GNOME
- You're the Boss with UBOS
- Multitenant Sites
- Linux for Astronomers