PUMA: 面向内存内处理架构的高效低成本内存分配与对齐支持
PUMA: Efficient and Low-Cost Memory Allocation and Alignment Support for Processing-Using-Memory Architectures
AI总结:
针对标准内存分配无法满足DRAM内存内处理架构的数据布局与对齐要求的问题,提出名为PUMA的内核级惰性分配例程,结合DRAM映射与大页拆分实现合规分配,实验验证其性能显著优于基线分配器。
摘要:
DRAM内存内处理(Processing-using-DRAM,PUD)架构对其操作数施加了严格的数据布局与对齐要求:源操作数和目的操作数(i)必须驻留在同一个DRAM子阵列中(即共享同一行缓冲器和行解码器的一组DRAM行),且(ii)需对齐到DRAM行的边界。然而,标准内存分配例程(即malloc、posix_memalign以及基于大页的内存分配)无法满足PUD架构正常运行所需的数据布局与对齐要求。为了让内存分配API能够影响操作系统内存分配器,并确保内存对象被放置在特定的DRAM子阵列内,我们在内核中提出了一种面向PUD内存对象的新型惰性数据分配例程,名为PUMA。PUMA的核心思路是结合DRAM内部映射信息与大页技术,随后将大页拆分为更细粒度的分配单元,这些单元(i)按页地址和大小对齐,且(ii)在虚拟地址空间中连续。\n我们使用QEMU将PUMA实现为一个内核模块,并模拟了一台运行Fedora 33系统、搭载v5.9.0版本Linux内核的RISC-V机器。我们模拟了一套PUD系统的实现,该系统能够执行行复制操作(如RowClone中的实现)以及布尔与/或/非运算(如Ambit中的实现)。在我们的实验中,若给定操作因数据未对齐而无法在PUD底层架构中执行,则会在主机CPU上执行该操作。在所有评估的微基准测试和分配规模下,PUMA的性能都显著优于基线内存分配器。
英文摘要:
Processing-using-DRAM (PUD) architectures impose a restrictive data layout and alignment for their operands, where source and destination operands (i) must reside in the same DRAM subarray (i.e., a group of DRAM rows sharing the same row buffer and row decoder) and (ii) are aligned to the boundaries of a DRAM row. However, standard memory allocation routines (i.e., malloc, posix_memalign, and huge pages-based memory allocation) fail to meet the data layout and alignment requirements for PUD architectures to operate successfully. To allow the memory allocation API to influence the OS memory allocator and ensure that memory objects are placed within specific DRAM subarrays, we propose a new lazy data allocation routine (in the kernel) for PUD memory objects called PUMA. The key idea of PUMA is to use the internal DRAM mapping information together with huge pages and then split huge pages into finer-grained allocation units that are (i) aligned to the page address and size and (ii) virtually contiguous. We implement PUMA as a kernel module using QEMU and emulate a RISC-V machine running Fedora 33 with v5.9.0 Linux Kernel. We emulate the implementation of a PUD system capable of executing row copy operations (as in RowClone) and Boolean AND/OR/NOT operations (as in Ambit). In our experiments, such an operation is performed in the host CPU if a given operation cannot be executed in our PUD substrate (due to data misalignment). PUMA significantly outperforms the baseline memory allocators for all evaluated microbenchmarks and allocation sizes.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。