This commit is a large set of code changes and results in removal of
hundreds, perhaps thousands, of CPU instructions, a portion of which
are on hot code paths.
Most (buffer *) used by lighttpd are not NULL, especially since buffers
were inlined into numerous larger structs such as request_st and chunk.
In the small number of instances where that is not the case, a NULL
check is often performed earlier in a function where that buffer is
later used with a buffer_* func. In the handful of cases that remained,
a NULL check was added, e.g. with r->http_host and r->conf.server_tag.
- check for empty strings at config time and set value to NULL if blank
string will be ignored at runtime; at runtime, simple pointer check
for NULL can be used to check for a value that has been set and is not
blank ("")
- use buffer_is_blank() instead of buffer_string_is_empty(),
and use buffer_is_unset() instead of buffer_is_empty(),
where buffer is known not to be NULL so that NULL check can be skipped
- use buffer_clen() instead of buffer_string_length() when buffer is
known not to be NULL (to avoid NULL check at runtime)
- use buffer_truncate() instead of buffer_string_set_length() to
truncate string, and use buffer_extend() to extend
Examples where buffer known not to be NULL:
- cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL
(though we might set it to NULL if buffer_is_blank(cpv->v.b))
- address of buffer is arg (&foo)
(compiler optimizer detects this in most, but not all, cases)
- buffer is checked for NULL earlier in func
- buffer is accessed in same scope without a NULL check (e.g. b->ptr)
internal behavior change:
callers must not pass a NULL buffer to some funcs.
- buffer_init_buffer() requires non-null args
- buffer_copy_buffer() requires non-null args
- buffer_append_string_buffer() requires non-null args
- buffer_string_space() requires non-null arg
move native data_* types into array.c
(the types are already declared in array.h)
The array data structure remains extendable, as is done with data_config
(configfile) and data_auth (mod_auth), though array data structure
primary uses are at startup (config time) and header parsing. The
insertion logic into sorted list can be expensive for large lists,
so header parsing might choose a different data structure in the future.
fix array_copy_array() sorted[]
(current use appears to be only on value lists; not user-visible issue)
Some time back, sorted[] was a set of indexes into data[],
and therefore sorted[] could be copied when copying array.
Now that sorted[] is a list of pointers, the sorted[] list
must be rebuilt. If copied, it continued to point to
(data_unset *) from the original src array, not the dst array.
(optional addition to (data_string *), used by http_header.[ch])
extend (data_string *) instead of creating another data_* TYPE_*
(new data type would probably have (data_string *) as base class)
(might revisit choice in the future)
HTTP_HEADER_UNSPECIFIED has been removed. It was used in select
locations as an optimization to avoid looking up enum header_header_e
before checking the array, but the ordering in the array now relies
on having the id. Having the id allows for a quick check if a common
header is present or not in the htags bitmask, before checking the
array, and allows for integer comparison in the log(n) search of the
array, instead of strncasecmp().
With HTTP_HEADER_UNSPECIFIED removed, add optimization to set bit
in htags for HTTP_HEADER_OTHER when an "other" header is added,
but do not clear the bit, as there might be addtl "other" headers
(expansion of buffer_string_lenth() inline function and CONST_BUF_LEN()
macro, which always check for NULL, appears to cause the analyzer to
believe that a pointer might be NULL in cases where it otherwise can
not be NULL)
x-ref:
http://clang-analyzer.llvm.org/faq.html
use global rather than passing around (server *) just for that
li_itostrn() and li_utostrn() return string length
(rather than requiring subsequent strlen() to find length)
convert all log_error_write() to log_error() and pass (log_error_st *)
use con->errh in preference to srv->errh (even though currently same)
avoid passing (server *) when previously used only for logging (errh)
While slightly more memory use in 64-bit (though same memory use as
prior versions of lighttpd), avoids bouncing through second array
when searching in sorted list. Most use of arrays in lighttpd is to
build a list once, and elements are not removed from the list.
This reverts commit 2260a8062ee599ecf28d9b52b981603fd2084aff.
original ordering of array elements is significant
e.g. in lighttpd.conf lists where first match to request is applied
return (data_unset *) from array_get_element_klen() to prep for
putting buffer into data_string for value member, rather than as
(buffer *)
(allow in-place modification of these buffer values)
array_get_element_klen() is now intended for read-only access
array_get_data_unset() is used by config processing for r/w access
array_get_buf_ptr() is used for r/w access to ds->value (string buffer)
short-circuit path to clear con->request.headers if entire size of
all request headers is <= BUFFER_MAX_REUSE_SIZE
clear (reset) data_string key and value upon reuse
mark array_get_index() as hot, rewrite to be pure and return sorted pos
mark routines as pure, as appropriate
mark routines as cold if used only at startup for config processing
mark params const, as appropriate
array_get_buf_ptr() for modifiable value buffer after insert into array
uint32_t used and size members instead of size_t
remove a->unique_ndx member; simply add to end of array for value lists
remove du->is_index_key member; simply check buffer_is_empty(du->key)
array_insert_key_value() used to be a hint that lookup could be skipped,
but the state from array_get_index() is now saved and reused internally,
so the distinction is no longer needed. Use array_set_key_value().
specialized buffer_eq_icase_ssn func replace strncasecmp()
in cases where string lengths are known to be at least as
large as the len being compared case-insensitively
regression in mod_evhost in lighttpd 1.4.53
regression in mod_flv_streaming in lighttpd 1.4.51 - lighttpd 1.4.53
(thx moisseev)
x-ref:
"[regression] %0 pattern does not match hostnames without the domain part"
https://redmine.lighttpd.net/issues/2932
reduce code duplication in list resizing
realloc() of NULL ptr has behavior similar to malloc()
Note that if initial size == 0, then code used to adjust size
must be += x to ensure the size is non-zero for reallocation.
(Multiplying 0 * x, e.g. power-2 resizing, will result in 0.)
save 40 bytes (64-bit), or 16 bytes (32-bit) per data_* element
at the cost of going through indirect function pointer to execute
methods. At runtime, the reset() method is most used among them.
More specific checks on contents of array lists. Each module using
lists now does better checking on the types of values in the list
(strings, integers, arrays/lists)
This helps prevent misconfiguration of things like cgi.assign,
fastcgi.server, and scgi.server, where source code might be
served as static files if parenthesis are misplaced.
x-ref:
https://redmine.lighttpd.net/boards/2/topics/6571
Glenn Strauss
Stefan Bühler