stream-clone: disable gc for the initial section for the v3 format
The number of small container created turn Python in a gc-frenzy that seriously
impact performance.
This significantly boost performance. The following number comes from a large
private repository using perf::stream-locked-section:
base-line: 35.04 seconds
prev-change: 24.51 seconds (-30%)
prev-change: 20.88 seconds (-40%)
this-change: 14.22 seconds (-60% from baseline; -31% from prev)
/*
* Utility functions
*
* Copyright (c) 2011 Yuya Nishihara <yuya@tcha.org>
*
* This software may be used and distributed according to the terms of the
* GNU General Public License version 2 or any later version.
*/
#ifndef UTIL_H_
#define UTIL_H_
#ifdef __GNUC__
#define PRINTF_FORMAT_ __attribute__((format(printf, 1, 2)))
#define UNUSED_ __attribute__((unused))
#else
#define PRINTF_FORMAT_
#define UNUSED_
#endif
void abortmsg(const char *fmt, ...) PRINTF_FORMAT_;
void abortmsgerrno(const char *fmt, ...) PRINTF_FORMAT_;
void enablecolor(void);
void enabledebugmsg(void);
void debugmsg(const char *fmt, ...) PRINTF_FORMAT_;
void fchdirx(int dirfd);
void fsetcloexec(int fd);
void *mallocx(size_t size);
void *reallocx(void *ptr, size_t size);
int runshellcmd(const char *cmd, const char *envp[], const char *cwd);
#endif /* UTIL_H_ */