| Name |
Last commit
|
Last update |
|---|---|---|
| bloat | ||
| crosstest | ||
| pydir | ||
| runtime | ||
| src | ||
| tests_lit | ||
| unittest | ||
| .gitignore | ||
| ALLOCATION.rst | ||
| LICENSE.TXT | ||
| LOWERING.rst | ||
| Makefile | ||
| Makefile.standalone | ||
| OWNERS | ||
| README.rst | ||
| codereview.settings |
The default LLVM raw_fd_ostream buffer size is based on stat'ing the FD and then checking st_blksize. Unfortunately, in the NaCl sandboxed build of pnacl-sz, NaCl's syscall returns st_blksize of 0 which makes the writes unbuffered. There is a comment in "src/trusted/service_runtime/include/bits/stat.h": nacl_abi_blksize_t nacl_abi_st_blksize; /* not implemented */ And the " src/trusted/desc/" implementation sets this to 0. This results in half a million write syscalls to translate the GCC pexe, which roughly doubles the translation time in sandboxed mode vs unsandboxed mode. Manually set a buffer size (Linux st_blksize seems to be about 4KB for comparison). This drops the number of write syscalls to about 200 for translating the GCC pexe. BUG= https://code.google.com/p/nativeclient/issues/detail?id=4091 R=stichnot@chromium.org Review URL: https://codereview.chromium.org/969403003
| Name |
Last commit
|
Last update |
|---|---|---|
| bloat | Loading commit data... | |
| crosstest | Loading commit data... | |
| pydir | Loading commit data... | |
| runtime | Loading commit data... | |
| src | Loading commit data... | |
| tests_lit | Loading commit data... | |
| unittest | Loading commit data... | |
| .gitignore | Loading commit data... | |
| ALLOCATION.rst | Loading commit data... | |
| LICENSE.TXT | Loading commit data... | |
| LOWERING.rst | Loading commit data... | |
| Makefile | Loading commit data... | |
| Makefile.standalone | Loading commit data... | |
| OWNERS | Loading commit data... | |
| README.rst | Loading commit data... | |
| codereview.settings | Loading commit data... |