mirror of
https://github.com/neovim/neovim.git
synced 2025-10-15 14:26:07 +00:00
refactor(grid): change schar_T representation to be more compact
Previously, a screen cell would occupy 28+4=32 bytes per cell as we always made space for up to MAX_MCO+1 codepoints in a cell. As an example, even a pretty modest 50*80 screen would consume 50*80*2*32 = 256000, i e a quarter megabyte With the factor of two due to the TUI side buffer, and even more when using msg_grid and/or ext_multigrid. This instead stores a 4-byte union of either: - a valid UTF-8 sequence up to 4 bytes - an escape char which is invalid UTF-8 (0xFF) plus a 24-bit index to a glyph cache This avoids allocating space for huge composed glyphs _upfront_, while still keeping rendering such glyphs reasonably fast (1 hash table lookup + one plain index lookup). If the same large glyphs are using repeatedly on the screen, this is still a net reduction of memory/cache consumption. The only case which really gets worse is if you blast the screen full with crazy emojis and zalgo text and even this case only leads to 4 extra bytes per char. When only <= 4-byte glyphs are used, plus the 4-byte attribute code, i e 8 bytes in total there is a factor of four reduction of memory use. Memory which will be quite hot in cache as the screen buffer is scanned over in win_line() buffer text drawing A slight complication is that the representation depends on host byte order. I've tested this manually by compling and running this in qemu-s390x and it works fine. We might add a qemu based solution to CI at some point.
This commit is contained in:
102
src/nvim/map_glyph_cache.c
Normal file
102
src/nvim/map_glyph_cache.c
Normal file
@@ -0,0 +1,102 @@
|
||||
// This is an open source non-commercial project. Dear PVS-Studio, please check
|
||||
// it. PVS-Studio Static Code Analyzer for C, C++ and C#: http://www.viva64.com
|
||||
|
||||
// Specialized version of Set() where interned strings is stored in a compact,
|
||||
// NUL-separated char array.
|
||||
// `String key` lookup keys don't need to be NULL terminated, but they
|
||||
// must not contain embedded NUL:s. When reading a key from set->keys, they
|
||||
// are always NUL terminated, though. Thus, it is enough to store an index into
|
||||
// this array, and use strlen(), to retrive an interned key.
|
||||
|
||||
#include "nvim/api/private/helpers.h"
|
||||
#include "nvim/map.h"
|
||||
|
||||
uint32_t mh_find_bucket_glyph(Set(glyph) *set, String key, bool put)
|
||||
{
|
||||
MapHash *h = &set->h;
|
||||
uint32_t step = 0;
|
||||
uint32_t mask = h->n_buckets - 1;
|
||||
uint32_t k = hash_String(key);
|
||||
uint32_t i = k & mask;
|
||||
uint32_t last = i;
|
||||
uint32_t site = put ? last : MH_TOMBSTONE;
|
||||
while (!mh_is_empty(h, i)) {
|
||||
if (mh_is_del(h, i)) {
|
||||
if (site == last) {
|
||||
site = i;
|
||||
}
|
||||
} else if (equal_String(cstr_as_string(&set->keys[h->hash[i] - 1]), key)) {
|
||||
return i;
|
||||
}
|
||||
i = (i + (++step)) & mask;
|
||||
if (i == last) {
|
||||
abort();
|
||||
}
|
||||
}
|
||||
if (site == last) {
|
||||
site = i;
|
||||
}
|
||||
return site;
|
||||
}
|
||||
|
||||
/// @return index into set->keys if found, MH_TOMBSTONE otherwise
|
||||
uint32_t mh_get_glyph(Set(glyph) *set, String key)
|
||||
{
|
||||
if (set->h.n_buckets == 0) {
|
||||
return MH_TOMBSTONE;
|
||||
}
|
||||
uint32_t idx = mh_find_bucket_glyph(set, key, false);
|
||||
return (idx != MH_TOMBSTONE) ? set->h.hash[idx] - 1 : MH_TOMBSTONE;
|
||||
}
|
||||
|
||||
void mh_rehash_glyph(Set(glyph) *set)
|
||||
{
|
||||
// assume the format of set->keys, i e NUL terminated strings
|
||||
for (uint32_t k = 0; k < set->h.n_keys; k += (uint32_t)strlen(&set->keys[k]) + 1) {
|
||||
uint32_t idx = mh_find_bucket_glyph(set, cstr_as_string(&set->keys[k]), true);
|
||||
// there must be tombstones when we do a rehash
|
||||
if (!mh_is_empty((&set->h), idx)) {
|
||||
abort();
|
||||
}
|
||||
set->h.hash[idx] = k + 1;
|
||||
}
|
||||
set->h.n_occupied = set->h.size = set->h.n_keys;
|
||||
}
|
||||
|
||||
uint32_t mh_put_glyph(Set(glyph) *set, String key, MHPutStatus *new)
|
||||
{
|
||||
MapHash *h = &set->h;
|
||||
// Might rehash ahead of time if "key" already existed. But it was
|
||||
// going to happen soon anyway.
|
||||
if (h->n_occupied >= h->upper_bound) {
|
||||
mh_realloc(h, h->n_buckets + 1);
|
||||
mh_rehash_glyph(set);
|
||||
}
|
||||
|
||||
uint32_t idx = mh_find_bucket_glyph(set, key, true);
|
||||
|
||||
if (mh_is_either(h, idx)) {
|
||||
h->size++;
|
||||
h->n_occupied++;
|
||||
|
||||
uint32_t size = (uint32_t)key.size + 1; // NUL takes space
|
||||
uint32_t pos = h->n_keys;
|
||||
h->n_keys += size;
|
||||
if (h->n_keys > h->keys_capacity) {
|
||||
h->keys_capacity = MAX(h->keys_capacity * 2, 64);
|
||||
set->keys = xrealloc(set->keys, h->keys_capacity * sizeof(char));
|
||||
*new = kMHNewKeyRealloc;
|
||||
} else {
|
||||
*new = kMHNewKeyDidFit;
|
||||
}
|
||||
memcpy(&set->keys[pos], key.data, key.size);
|
||||
set->keys[pos + key.size] = NUL;
|
||||
h->hash[idx] = pos + 1;
|
||||
return pos;
|
||||
} else {
|
||||
*new = kMHExisting;
|
||||
uint32_t pos = h->hash[idx] - 1;
|
||||
assert(equal_String(cstr_as_string(&set->keys[pos]), key));
|
||||
return pos;
|
||||
}
|
||||
}
|
Reference in New Issue
Block a user