^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (C) 2018 HUAWEI, Inc.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) * https://www.huawei.com/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Created by Gao Xiang <gaoxiang25@huawei.com>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #include "zdata.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) #include "compress.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include <linux/prefetch.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include <trace/events/erofs.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * since pclustersize is variable for big pcluster feature, introduce slab
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * pools implementation for different pcluster sizes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) struct z_erofs_pcluster_slab {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) struct kmem_cache *slab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) unsigned int maxpages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) char name[48];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) #define _PCLP(n) { .maxpages = n }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) static struct z_erofs_pcluster_slab pcluster_pool[] __read_mostly = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) _PCLP(1), _PCLP(4), _PCLP(16), _PCLP(64), _PCLP(128),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) _PCLP(Z_EROFS_PCLUSTER_MAX_PAGES)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) static void z_erofs_destroy_pcluster_pool(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) for (i = 0; i < ARRAY_SIZE(pcluster_pool); ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) if (!pcluster_pool[i].slab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) kmem_cache_destroy(pcluster_pool[i].slab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) pcluster_pool[i].slab = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) static int z_erofs_create_pcluster_pool(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) struct z_erofs_pcluster_slab *pcs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) struct z_erofs_pcluster *a;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) unsigned int size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) for (pcs = pcluster_pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) pcs < pcluster_pool + ARRAY_SIZE(pcluster_pool); ++pcs) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) size = struct_size(a, compressed_pages, pcs->maxpages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) sprintf(pcs->name, "erofs_pcluster-%u", pcs->maxpages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) pcs->slab = kmem_cache_create(pcs->name, size, 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) SLAB_RECLAIM_ACCOUNT, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) if (pcs->slab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) z_erofs_destroy_pcluster_pool();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) static struct z_erofs_pcluster *z_erofs_alloc_pcluster(unsigned int nrpages)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) for (i = 0; i < ARRAY_SIZE(pcluster_pool); ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) struct z_erofs_pcluster_slab *pcs = pcluster_pool + i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) struct z_erofs_pcluster *pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) if (nrpages > pcs->maxpages)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) pcl = kmem_cache_zalloc(pcs->slab, GFP_NOFS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) if (!pcl)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) return ERR_PTR(-ENOMEM);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) pcl->pclusterpages = nrpages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) return pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) return ERR_PTR(-EINVAL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) static void z_erofs_free_pcluster(struct z_erofs_pcluster *pcl)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) for (i = 0; i < ARRAY_SIZE(pcluster_pool); ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) struct z_erofs_pcluster_slab *pcs = pcluster_pool + i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) if (pcl->pclusterpages > pcs->maxpages)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) kmem_cache_free(pcs->slab, pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) * a compressed_pages[] placeholder in order to avoid
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) * being filled with file pages for in-place decompression.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) #define PAGE_UNALLOCATED ((void *)0x5F0E4B1D)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) /* how to allocate cached pages for a pcluster */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) enum z_erofs_cache_alloctype {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) DONTALLOC, /* don't allocate any cached pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) DELAYEDALLOC, /* delayed allocation (at the time of submitting io) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) * try to use cached I/O if page allocation succeeds or fallback
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) * to in-place I/O instead to avoid any direct reclaim.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) TRYALLOC,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) * tagged pointer with 1-bit tag for all compressed pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) * tag 0 - the page is just found with an extra page reference
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) typedef tagptr1_t compressed_page_t;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) #define tag_compressed_page_justfound(page) \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) tagptr_fold(compressed_page_t, page, 1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) static struct workqueue_struct *z_erofs_workqueue __read_mostly;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) void z_erofs_exit_zip_subsystem(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) destroy_workqueue(z_erofs_workqueue);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) z_erofs_destroy_pcluster_pool();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) static inline int z_erofs_init_workqueue(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) const unsigned int onlinecpus = num_possible_cpus();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) * no need to spawn too many threads, limiting threads could minimum
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) * scheduling overhead, perhaps per-CPU threads should be better?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) z_erofs_workqueue = alloc_workqueue("erofs_unzipd",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) WQ_UNBOUND | WQ_HIGHPRI,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) onlinecpus + onlinecpus / 4);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) return z_erofs_workqueue ? 0 : -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) int __init z_erofs_init_zip_subsystem(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) int err = z_erofs_create_pcluster_pool();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) err = z_erofs_init_workqueue();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) z_erofs_destroy_pcluster_pool();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) enum z_erofs_collectmode {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) COLLECT_SECONDARY,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) COLLECT_PRIMARY,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) * The current collection was the tail of an exist chain, in addition
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) * that the previous processed chained collections are all decided to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) * be hooked up to it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) * A new chain will be created for the remaining collections which are
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) * not processed yet, therefore different from COLLECT_PRIMARY_FOLLOWED,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) * the next collection cannot reuse the whole page safely in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) * the following scenario:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) * ________________________________________________________________
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) * | tail (partial) page | head (partial) page |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) * | (belongs to the next cl) | (belongs to the current cl) |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) * |_______PRIMARY_FOLLOWED_______|________PRIMARY_HOOKED___________|
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) COLLECT_PRIMARY_HOOKED,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) * a weak form of COLLECT_PRIMARY_FOLLOWED, the difference is that it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) * could be dispatched into bypass queue later due to uptodated managed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) * pages. All related online pages cannot be reused for inplace I/O (or
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) * pagevec) since it can be directly decoded without I/O submission.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) COLLECT_PRIMARY_FOLLOWED_NOINPLACE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) * The current collection has been linked with the owned chain, and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) * could also be linked with the remaining collections, which means
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) * if the processing page is the tail page of the collection, thus
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) * the current collection can safely use the whole page (since
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) * the previous collection is under control) for in-place I/O, as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) * illustrated below:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) * ________________________________________________________________
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) * | tail (partial) page | head (partial) page |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) * | (of the current cl) | (of the previous collection) |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) * | PRIMARY_FOLLOWED or | |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) * |_____PRIMARY_HOOKED___|____________PRIMARY_FOLLOWED____________|
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) * [ (*) the above page can be used as inplace I/O. ]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) COLLECT_PRIMARY_FOLLOWED,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) struct z_erofs_collector {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) struct z_erofs_pagevec_ctor vector;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) struct z_erofs_pcluster *pcl, *tailpcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) struct z_erofs_collection *cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) /* a pointer used to pick up inplace I/O pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) struct page **icpage_ptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) z_erofs_next_pcluster_t owned_head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) enum z_erofs_collectmode mode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) struct z_erofs_decompress_frontend {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) struct inode *const inode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) struct z_erofs_collector clt;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) struct erofs_map_blocks map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) bool readahead;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) /* used for applying cache strategy on the fly */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) bool backmost;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) erofs_off_t headoffset;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) #define COLLECTOR_INIT() { \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) .owned_head = Z_EROFS_PCLUSTER_TAIL, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) .mode = COLLECT_PRIMARY_FOLLOWED }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) #define DECOMPRESS_FRONTEND_INIT(__i) { \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) .inode = __i, .clt = COLLECTOR_INIT(), \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) .backmost = true, }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) static struct page *z_pagemap_global[Z_EROFS_VMAP_GLOBAL_PAGES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) static DEFINE_MUTEX(z_pagemap_global_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) static void preload_compressed_pages(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) struct address_space *mc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) enum z_erofs_cache_alloctype type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) struct list_head *pagepool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) struct z_erofs_pcluster *pcl = clt->pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) bool standalone = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) gfp_t gfp = (mapping_gfp_mask(mc) & ~__GFP_DIRECT_RECLAIM) |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) struct page **pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) pgoff_t index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) if (clt->mode < COLLECT_PRIMARY_FOLLOWED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) pages = pcl->compressed_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) index = pcl->obj.index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) for (; index < pcl->obj.index + pcl->pclusterpages; ++index, ++pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) compressed_page_t t;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) struct page *newpage = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) /* the compressed page was loaded before */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) if (READ_ONCE(*pages))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) page = find_get_page(mc, index);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) if (page) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) t = tag_compressed_page_justfound(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) /* I/O is needed, no possible to decompress directly */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) standalone = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) switch (type) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) case DELAYEDALLOC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) t = tagptr_init(compressed_page_t,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) PAGE_UNALLOCATED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) case TRYALLOC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) newpage = erofs_allocpage(pagepool, gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) if (!newpage)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) set_page_private(newpage,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) Z_EROFS_PREALLOCATED_PAGE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) t = tag_compressed_page_justfound(newpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) default: /* DONTALLOC */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) if (!cmpxchg_relaxed(pages, NULL, tagptr_cast_ptr(t)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) if (page) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) put_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) } else if (newpage) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) set_page_private(newpage, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) list_add(&newpage->lru, pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) * don't do inplace I/O if all compressed pages are available in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) * managed cache since it can be moved to the bypass queue instead.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) if (standalone)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) clt->mode = COLLECT_PRIMARY_FOLLOWED_NOINPLACE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) /* called by erofs_shrinker to get rid of all compressed_pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) int erofs_try_to_free_all_cached_pages(struct erofs_sb_info *sbi,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) struct erofs_workgroup *grp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) struct z_erofs_pcluster *const pcl =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) container_of(grp, struct z_erofs_pcluster, obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) struct address_space *const mapping = MNGD_MAPPING(sbi);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) * refcount of workgroup is now freezed as 1,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) * therefore no need to worry about available decompression users.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) for (i = 0; i < pcl->pclusterpages; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) struct page *page = pcl->compressed_pages[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) /* block other users from reclaiming or migrating the page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) if (!trylock_page(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) if (page->mapping != mapping)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) /* barrier is implied in the following 'unlock_page' */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) WRITE_ONCE(pcl->compressed_pages[i], NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) detach_page_private(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) int erofs_try_to_free_cached_page(struct address_space *mapping,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) struct z_erofs_pcluster *const pcl = (void *)page_private(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) int ret = 0; /* 0 - busy */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) if (erofs_workgroup_try_to_freeze(&pcl->obj, 1)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) unsigned int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) for (i = 0; i < pcl->pclusterpages; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) if (pcl->compressed_pages[i] == page) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) WRITE_ONCE(pcl->compressed_pages[i], NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) ret = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) erofs_workgroup_unfreeze(&pcl->obj, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) detach_page_private(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) /* page_type must be Z_EROFS_PAGE_TYPE_EXCLUSIVE */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) static bool z_erofs_try_inplace_io(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) struct z_erofs_pcluster *const pcl = clt->pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) while (clt->icpage_ptr > pcl->compressed_pages)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) if (!cmpxchg(--clt->icpage_ptr, NULL, page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) /* callers must be with collection lock held */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) static int z_erofs_attach_page(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) struct page *page, enum z_erofs_page_type type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) bool pvec_safereuse)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) /* give priority for inplaceio */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) if (clt->mode >= COLLECT_PRIMARY &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) type == Z_EROFS_PAGE_TYPE_EXCLUSIVE &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) z_erofs_try_inplace_io(clt, page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) ret = z_erofs_pagevec_enqueue(&clt->vector, page, type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) pvec_safereuse);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) clt->cl->vcnt += (unsigned int)ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) return ret ? 0 : -EAGAIN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) static enum z_erofs_collectmode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) try_to_claim_pcluster(struct z_erofs_pcluster *pcl,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) z_erofs_next_pcluster_t *owned_head)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) /* let's claim these following types of pclusters */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) retry:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) if (pcl->next == Z_EROFS_PCLUSTER_NIL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) /* type 1, nil pcluster */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) if (cmpxchg(&pcl->next, Z_EROFS_PCLUSTER_NIL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) *owned_head) != Z_EROFS_PCLUSTER_NIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) goto retry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) *owned_head = &pcl->next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) /* lucky, I am the followee :) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) return COLLECT_PRIMARY_FOLLOWED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) } else if (pcl->next == Z_EROFS_PCLUSTER_TAIL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) * type 2, link to the end of a existing open chain,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) * be careful that its submission itself is governed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) * by the original owned chain.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) if (cmpxchg(&pcl->next, Z_EROFS_PCLUSTER_TAIL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) *owned_head) != Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) goto retry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) *owned_head = Z_EROFS_PCLUSTER_TAIL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) return COLLECT_PRIMARY_HOOKED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) return COLLECT_PRIMARY; /* :( better luck next time */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) static int z_erofs_lookup_collection(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) struct inode *inode,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) struct erofs_map_blocks *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) struct z_erofs_pcluster *pcl = clt->pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) struct z_erofs_collection *cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) unsigned int length;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) /* to avoid unexpected loop formed by corrupted images */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) if (clt->owned_head == &pcl->next || pcl == clt->tailpcl) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) return -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) cl = z_erofs_primarycollection(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) if (cl->pageofs != (map->m_la & ~PAGE_MASK)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) return -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) length = READ_ONCE(pcl->length);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) if (length & Z_EROFS_PCLUSTER_FULL_LENGTH) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) if ((map->m_llen << Z_EROFS_PCLUSTER_LENGTH_BIT) > length) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) return -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) unsigned int llen = map->m_llen << Z_EROFS_PCLUSTER_LENGTH_BIT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) if (map->m_flags & EROFS_MAP_FULL_MAPPED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) llen |= Z_EROFS_PCLUSTER_FULL_LENGTH;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) while (llen > length &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) length != cmpxchg_relaxed(&pcl->length, length, llen)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) cpu_relax();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) length = READ_ONCE(pcl->length);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) mutex_lock(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) /* used to check tail merging loop due to corrupted images */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) if (clt->owned_head == Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) clt->tailpcl = pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) clt->mode = try_to_claim_pcluster(pcl, &clt->owned_head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) /* clean tailpcl if the current owned_head is Z_EROFS_PCLUSTER_TAIL */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) if (clt->owned_head == Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) clt->tailpcl = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) clt->cl = cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) static int z_erofs_register_collection(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) struct inode *inode,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) struct erofs_map_blocks *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) struct z_erofs_pcluster *pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) struct z_erofs_collection *cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) struct erofs_workgroup *grp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) int err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) /* no available pcluster, let's allocate one */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) pcl = z_erofs_alloc_pcluster(map->m_plen >> PAGE_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) if (IS_ERR(pcl))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) return PTR_ERR(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) atomic_set(&pcl->obj.refcount, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) pcl->obj.index = map->m_pa >> PAGE_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) pcl->length = (map->m_llen << Z_EROFS_PCLUSTER_LENGTH_BIT) |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) (map->m_flags & EROFS_MAP_FULL_MAPPED ?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) Z_EROFS_PCLUSTER_FULL_LENGTH : 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497) if (map->m_flags & EROFS_MAP_ZIPPED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) pcl->algorithmformat = Z_EROFS_COMPRESSION_LZ4;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) pcl->algorithmformat = Z_EROFS_COMPRESSION_SHIFTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) /* new pclusters should be claimed as type 1, primary and followed */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) pcl->next = clt->owned_head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) clt->mode = COLLECT_PRIMARY_FOLLOWED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) cl = z_erofs_primarycollection(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) cl->pageofs = map->m_la & ~PAGE_MASK;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) * lock all primary followed works before visible to others
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) * and mutex_trylock *never* fails for a new pcluster.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) mutex_init(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) DBG_BUGON(!mutex_trylock(&cl->lock));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) grp = erofs_insert_workgroup(inode->i_sb, &pcl->obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) if (IS_ERR(grp)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) err = PTR_ERR(grp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) goto err_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) if (grp != &pcl->obj) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) clt->pcl = container_of(grp, struct z_erofs_pcluster, obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) err = -EEXIST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) goto err_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) /* used to check tail merging loop due to corrupted images */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) if (clt->owned_head == Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) clt->tailpcl = pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) clt->owned_head = &pcl->next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) clt->pcl = pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532) clt->cl = cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) err_out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) mutex_unlock(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) z_erofs_free_pcluster(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) static int z_erofs_collector_begin(struct z_erofs_collector *clt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) struct inode *inode,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543) struct erofs_map_blocks *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545) struct erofs_workgroup *grp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) DBG_BUGON(clt->cl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) /* must be Z_EROFS_PCLUSTER_TAIL or pointed to previous collection */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) DBG_BUGON(clt->owned_head == Z_EROFS_PCLUSTER_NIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) DBG_BUGON(clt->owned_head == Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) if (!PAGE_ALIGNED(map->m_pa)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) grp = erofs_find_workgroup(inode->i_sb, map->m_pa >> PAGE_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) if (grp) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) clt->pcl = container_of(grp, struct z_erofs_pcluster, obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) ret = z_erofs_register_collection(clt, inode, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565) if (!ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) if (ret != -EEXIST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571) ret = z_erofs_lookup_collection(clt, inode, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) if (ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) erofs_workgroup_put(&clt->pcl->obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) z_erofs_pagevec_ctor_init(&clt->vector, Z_EROFS_NR_INLINE_PAGEVECS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) clt->cl->pagevec, clt->cl->vcnt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) /* since file-backed online pages are traversed in reverse order */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) clt->icpage_ptr = clt->pcl->compressed_pages + clt->pcl->pclusterpages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587) * keep in mind that no referenced pclusters will be freed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588) * only after a RCU grace period.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) static void z_erofs_rcu_callback(struct rcu_head *head)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) struct z_erofs_collection *const cl =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593) container_of(head, struct z_erofs_collection, rcu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595) z_erofs_free_pcluster(container_of(cl, struct z_erofs_pcluster,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596) primary_collection));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) void erofs_workgroup_free_rcu(struct erofs_workgroup *grp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) struct z_erofs_pcluster *const pcl =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) container_of(grp, struct z_erofs_pcluster, obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) struct z_erofs_collection *const cl = z_erofs_primarycollection(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) call_rcu(&cl->rcu, z_erofs_rcu_callback);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) static void z_erofs_collection_put(struct z_erofs_collection *cl)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) struct z_erofs_pcluster *const pcl =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) container_of(cl, struct z_erofs_pcluster, primary_collection);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) erofs_workgroup_put(&pcl->obj);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) static bool z_erofs_collector_end(struct z_erofs_collector *clt)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) struct z_erofs_collection *cl = clt->cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) if (!cl)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) z_erofs_pagevec_ctor_exit(&clt->vector, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) mutex_unlock(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627) * if all pending pages are added, don't hold its reference
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) * any longer if the pcluster isn't hosted by ourselves.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630) if (clt->mode < COLLECT_PRIMARY_FOLLOWED_NOINPLACE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) z_erofs_collection_put(cl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633) clt->cl = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) static bool should_alloc_managed_pages(struct z_erofs_decompress_frontend *fe,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) unsigned int cachestrategy,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639) erofs_off_t la)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641) if (cachestrategy <= EROFS_ZIP_CACHE_DISABLED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) if (fe->backmost)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) return cachestrategy >= EROFS_ZIP_CACHE_READAROUND &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648) la < fe->headoffset;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651) static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) struct page *page, struct list_head *pagepool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) struct inode *const inode = fe->inode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655) struct erofs_sb_info *const sbi = EROFS_I_SB(inode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) struct erofs_map_blocks *const map = &fe->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657) struct z_erofs_collector *const clt = &fe->clt;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658) const loff_t offset = page_offset(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) bool tight = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661) enum z_erofs_cache_alloctype cache_strategy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) enum z_erofs_page_type page_type;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) unsigned int cur, end, spiltted, index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) int err = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666) /* register locked file pages as online pages in pack */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667) z_erofs_onlinepage_init(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) spiltted = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) end = PAGE_SIZE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671) repeat:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) cur = end - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674) /* lucky, within the range of the current map_blocks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675) if (offset + cur >= map->m_la &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676) offset + cur < map->m_la + map->m_llen) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) /* didn't get a valid collection previously (very rare) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678) if (!clt->cl)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) goto restart_now;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) goto hitted;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) /* go ahead the next map_blocks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) erofs_dbg("%s: [out-of-range] pos %llu", __func__, offset + cur);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686) if (z_erofs_collector_end(clt))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687) fe->backmost = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) map->m_la = offset + cur;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690) map->m_llen = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691) err = z_erofs_map_blocks_iter(inode, map, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) goto err_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) restart_now:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696) if (!(map->m_flags & EROFS_MAP_MAPPED))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) goto hitted;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699) err = z_erofs_collector_begin(clt, inode, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) goto err_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) /* preload all compressed pages (maybe downgrade role if necessary) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704) if (should_alloc_managed_pages(fe, sbi->ctx.cache_strategy, map->m_la))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705) cache_strategy = TRYALLOC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707) cache_strategy = DONTALLOC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) preload_compressed_pages(clt, MNGD_MAPPING(sbi),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710) cache_strategy, pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712) hitted:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) * Ensure the current partial page belongs to this submit chain rather
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) * than other concurrent submit chains or the noio(bypass) chain since
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716) * those chains are handled asynchronously thus the page cannot be used
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) * for inplace I/O or pagevec (should be processed in strict order.)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719) tight &= (clt->mode >= COLLECT_PRIMARY_HOOKED &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720) clt->mode != COLLECT_PRIMARY_FOLLOWED_NOINPLACE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) cur = end - min_t(unsigned int, offset + end - map->m_la, end);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723) if (!(map->m_flags & EROFS_MAP_MAPPED)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) zero_user_segment(page, cur, end);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725) goto next_part;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) /* let's derive page type */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729) page_type = cur ? Z_EROFS_VLE_PAGE_TYPE_HEAD :
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730) (!spiltted ? Z_EROFS_PAGE_TYPE_EXCLUSIVE :
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731) (tight ? Z_EROFS_PAGE_TYPE_EXCLUSIVE :
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732) Z_EROFS_VLE_PAGE_TYPE_TAIL_SHARED));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) if (cur)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) tight &= (clt->mode >= COLLECT_PRIMARY_FOLLOWED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737) retry:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) err = z_erofs_attach_page(clt, page, page_type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739) clt->mode >= COLLECT_PRIMARY_FOLLOWED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) /* should allocate an additional staging page for pagevec */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741) if (err == -EAGAIN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) struct page *const newpage =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743) alloc_page(GFP_NOFS | __GFP_NOFAIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) set_page_private(newpage, Z_EROFS_SHORTLIVED_PAGE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) err = z_erofs_attach_page(clt, newpage,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747) Z_EROFS_PAGE_TYPE_EXCLUSIVE, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) if (!err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749) goto retry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) goto err_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) index = page->index - (map->m_la >> PAGE_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) z_erofs_onlinepage_fixup(page, index, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759) /* bump up the number of spiltted parts of a page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) ++spiltted;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761) /* also update nr_pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762) clt->cl->nr_pages = max_t(pgoff_t, clt->cl->nr_pages, index + 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) next_part:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764) /* can be used for verification */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) map->m_llen = offset + cur - map->m_la;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767) end = cur;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) if (end > 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769) goto repeat;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) z_erofs_onlinepage_endio(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774) erofs_dbg("%s, finish page: %pK spiltted: %u map->m_llen %llu",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775) __func__, page, spiltted, map->m_llen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778) /* if some error occurred while processing this page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779) err_out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) SetPageError(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784) static void z_erofs_decompressqueue_work(struct work_struct *work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785) static void z_erofs_decompress_kickoff(struct z_erofs_decompressqueue *io,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786) bool sync, int bios)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) struct erofs_sb_info *const sbi = EROFS_SB(io->sb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790) /* wake up the caller thread for sync decompression */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) if (sync) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794) spin_lock_irqsave(&io->u.wait.lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795) if (!atomic_add_return(bios, &io->pending_bios))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796) wake_up_locked(&io->u.wait);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) spin_unlock_irqrestore(&io->u.wait.lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) if (atomic_add_return(bios, &io->pending_bios))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) /* Use workqueue and sync decompression for atomic contexts only */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804) if (in_atomic() || irqs_disabled()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805) queue_work(z_erofs_workqueue, &io->u.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806) sbi->ctx.readahead_sync_decompress = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) z_erofs_decompressqueue_work(&io->u.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812) static bool z_erofs_page_is_invalidated(struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) return !page->mapping && !z_erofs_is_shortlived_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) static void z_erofs_decompressqueue_endio(struct bio *bio)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) tagptr1_t t = tagptr_init(tagptr1_t, bio->bi_private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820) struct z_erofs_decompressqueue *q = tagptr_unfold_ptr(t);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) blk_status_t err = bio->bi_status;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) struct bio_vec *bvec;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823) struct bvec_iter_all iter_all;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) bio_for_each_segment_all(bvec, bio, iter_all) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826) struct page *page = bvec->bv_page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) DBG_BUGON(PageUptodate(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829) DBG_BUGON(z_erofs_page_is_invalidated(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832) SetPageError(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834) if (erofs_page_is_managed(EROFS_SB(q->sb), page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) if (!err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836) SetPageUptodate(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840) z_erofs_decompress_kickoff(q, tagptr_unfold_tags(t), -1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841) bio_put(bio);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) static int z_erofs_decompress_pcluster(struct super_block *sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) struct z_erofs_pcluster *pcl,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) struct list_head *pagepool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) struct erofs_sb_info *const sbi = EROFS_SB(sb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849) struct z_erofs_pagevec_ctor ctor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) unsigned int i, inputsize, outputsize, llen, nr_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851) struct page *pages_onstack[Z_EROFS_VMAP_ONSTACK_PAGES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) struct page **pages, **compressed_pages, *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) enum z_erofs_page_type page_type;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855) bool overlapped, partial;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) struct z_erofs_collection *cl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) int err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859) might_sleep();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) cl = z_erofs_primarycollection(pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861) DBG_BUGON(!READ_ONCE(cl->nr_pages));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) mutex_lock(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864) nr_pages = cl->nr_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) if (nr_pages <= Z_EROFS_VMAP_ONSTACK_PAGES) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) pages = pages_onstack;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) } else if (nr_pages <= Z_EROFS_VMAP_GLOBAL_PAGES &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869) mutex_trylock(&z_pagemap_global_lock)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) pages = z_pagemap_global;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872) gfp_t gfp_flags = GFP_KERNEL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874) if (nr_pages > Z_EROFS_VMAP_GLOBAL_PAGES)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) gfp_flags |= __GFP_NOFAIL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877) pages = kvmalloc_array(nr_pages, sizeof(struct page *),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878) gfp_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880) /* fallback to global pagemap for the lowmem scenario */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881) if (!pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) mutex_lock(&z_pagemap_global_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) pages = z_pagemap_global;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887) for (i = 0; i < nr_pages; ++i)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) pages[i] = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890) err = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) z_erofs_pagevec_ctor_init(&ctor, Z_EROFS_NR_INLINE_PAGEVECS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892) cl->pagevec, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894) for (i = 0; i < cl->vcnt; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895) unsigned int pagenr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) page = z_erofs_pagevec_dequeue(&ctor, &page_type);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) /* all pages in pagevec ought to be valid */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900) DBG_BUGON(!page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901) DBG_BUGON(z_erofs_page_is_invalidated(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903) if (z_erofs_put_shortlivedpage(pagepool, page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) if (page_type == Z_EROFS_VLE_PAGE_TYPE_HEAD)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907) pagenr = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) pagenr = z_erofs_onlinepage_index(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) DBG_BUGON(pagenr >= nr_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914) * currently EROFS doesn't support multiref(dedup),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) * so here erroring out one multiref page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) if (pages[pagenr]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919) SetPageError(pages[pagenr]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) z_erofs_onlinepage_endio(pages[pagenr]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921) err = -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) pages[pagenr] = page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) z_erofs_pagevec_ctor_exit(&ctor, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) overlapped = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928) compressed_pages = pcl->compressed_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930) for (i = 0; i < pcl->pclusterpages; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931) unsigned int pagenr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) page = compressed_pages[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935) /* all compressed pages ought to be valid */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936) DBG_BUGON(!page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937) DBG_BUGON(z_erofs_page_is_invalidated(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939) if (!z_erofs_is_shortlived_page(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940) if (erofs_page_is_managed(sbi, page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) if (!PageUptodate(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) err = -EIO;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) * only if non-head page can be selected
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948) * for inplace decompression
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950) pagenr = z_erofs_onlinepage_index(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) DBG_BUGON(pagenr >= nr_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953) if (pages[pagenr]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) DBG_BUGON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955) SetPageError(pages[pagenr]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956) z_erofs_onlinepage_endio(pages[pagenr]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) err = -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959) pages[pagenr] = page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961) overlapped = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) /* PG_error needs checking for all non-managed pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965) if (PageError(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) DBG_BUGON(PageUptodate(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) err = -EIO;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974) llen = pcl->length >> Z_EROFS_PCLUSTER_LENGTH_BIT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975) if (nr_pages << PAGE_SHIFT >= cl->pageofs + llen) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976) outputsize = llen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) partial = !(pcl->length & Z_EROFS_PCLUSTER_FULL_LENGTH);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979) outputsize = (nr_pages << PAGE_SHIFT) - cl->pageofs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) partial = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983) inputsize = pcl->pclusterpages * PAGE_SIZE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984) err = z_erofs_decompress(&(struct z_erofs_decompress_req) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) .sb = sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) .in = compressed_pages,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) .out = pages,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988) .pageofs_out = cl->pageofs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) .inputsize = inputsize,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990) .outputsize = outputsize,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) .alg = pcl->algorithmformat,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992) .inplace_io = overlapped,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993) .partial_decoding = partial
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994) }, pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997) /* must handle all compressed pages before ending pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) for (i = 0; i < pcl->pclusterpages; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999) page = compressed_pages[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001) if (erofs_page_is_managed(sbi, page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004) /* recycle all individual short-lived pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) (void)z_erofs_put_shortlivedpage(pagepool, page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) WRITE_ONCE(compressed_pages[i], NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) for (i = 0; i < nr_pages; ++i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) page = pages[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) DBG_BUGON(z_erofs_page_is_invalidated(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017) /* recycle all individual short-lived pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) if (z_erofs_put_shortlivedpage(pagepool, page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) if (err < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022) SetPageError(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) z_erofs_onlinepage_endio(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027) if (pages == z_pagemap_global)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028) mutex_unlock(&z_pagemap_global_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029) else if (pages != pages_onstack)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030) kvfree(pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032) cl->nr_pages = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) cl->vcnt = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035) /* all cl locks MUST be taken before the following line */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036) WRITE_ONCE(pcl->next, Z_EROFS_PCLUSTER_NIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038) /* all cl locks SHOULD be released right now */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) mutex_unlock(&cl->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041) z_erofs_collection_put(cl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045) static void z_erofs_decompress_queue(const struct z_erofs_decompressqueue *io,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) struct list_head *pagepool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048) z_erofs_next_pcluster_t owned = io->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050) while (owned != Z_EROFS_PCLUSTER_TAIL_CLOSED) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) struct z_erofs_pcluster *pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053) /* no possible that 'owned' equals Z_EROFS_WORK_TPTR_TAIL */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) DBG_BUGON(owned == Z_EROFS_PCLUSTER_TAIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056) /* no possible that 'owned' equals NULL */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057) DBG_BUGON(owned == Z_EROFS_PCLUSTER_NIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) pcl = container_of(owned, struct z_erofs_pcluster, next);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060) owned = READ_ONCE(pcl->next);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062) z_erofs_decompress_pcluster(io->sb, pcl, pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066) static void z_erofs_decompressqueue_work(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068) struct z_erofs_decompressqueue *bgq =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069) container_of(work, struct z_erofs_decompressqueue, u.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) LIST_HEAD(pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) DBG_BUGON(bgq->head == Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073) z_erofs_decompress_queue(bgq, &pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) put_pages_list(&pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) kvfree(bgq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079) static struct page *pickup_page_for_submission(struct z_erofs_pcluster *pcl,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080) unsigned int nr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081) struct list_head *pagepool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) struct address_space *mc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083) gfp_t gfp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) const pgoff_t index = pcl->obj.index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086) bool tocache = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088) struct address_space *mapping;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) struct page *oldpage, *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091) compressed_page_t t;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092) int justfound;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) repeat:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095) page = READ_ONCE(pcl->compressed_pages[nr]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) oldpage = page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) goto out_allocpage;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102) * the cached page has not been allocated and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103) * an placeholder is out there, prepare it now.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105) if (page == PAGE_UNALLOCATED) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106) tocache = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) goto out_allocpage;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) /* process the target tagged pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111) t = tagptr_init(compressed_page_t, page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) justfound = tagptr_unfold_tags(t);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) page = tagptr_unfold_ptr(t);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) * preallocated cached pages, which is used to avoid direct reclaim
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) * otherwise, it will go inplace I/O path instead.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) if (page->private == Z_EROFS_PREALLOCATED_PAGE) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120) WRITE_ONCE(pcl->compressed_pages[nr], page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121) set_page_private(page, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) tocache = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) goto out_tocache;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) mapping = READ_ONCE(page->mapping);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128) * file-backed online pages in plcuster are all locked steady,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) * therefore it is impossible for `mapping' to be NULL.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131) if (mapping && mapping != mc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132) /* ought to be unmanaged pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135) /* directly return for shortlived page as well */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136) if (z_erofs_is_shortlived_page(page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) lock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141) /* only true if page reclaim goes wrong, should never happen */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142) DBG_BUGON(justfound && PagePrivate(page));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144) /* the page is still in manage cache */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145) if (page->mapping == mc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) WRITE_ONCE(pcl->compressed_pages[nr], page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148) ClearPageError(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) if (!PagePrivate(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151) * impossible to be !PagePrivate(page) for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152) * the current restriction as well if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153) * the page is already in compressed_pages[].
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) DBG_BUGON(!justfound);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) justfound = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158) set_page_private(page, (unsigned long)pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) SetPagePrivate(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) /* no need to submit io if it is already up-to-date */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) if (PageUptodate(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165) page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) * the managed page has been truncated, it's unsafe to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172) * reuse this one, let's allocate a new cache-managed page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174) DBG_BUGON(page->mapping);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) DBG_BUGON(!justfound);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177) tocache = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) put_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180) out_allocpage:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) page = erofs_allocpage(pagepool, gfp | __GFP_NOFAIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182) if (oldpage != cmpxchg(&pcl->compressed_pages[nr], oldpage, page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) list_add(&page->lru, pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185) goto repeat;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187) out_tocache:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188) if (!tocache || add_to_page_cache_lru(page, mc, index + nr, gfp)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) /* turn into temporary page if fails (1 ref) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) set_page_private(page, Z_EROFS_SHORTLIVED_PAGE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193) attach_page_private(page, pcl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) /* drop a refcount added by allocpage (then we have 2 refs here) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1195) put_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1196)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1197) out: /* the only exit (for tracing and debugging) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1198) return page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1199) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1201) static struct z_erofs_decompressqueue *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1202) jobqueue_init(struct super_block *sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1203) struct z_erofs_decompressqueue *fgq, bool *fg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1204) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1205) struct z_erofs_decompressqueue *q;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1206)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1207) if (fg && !*fg) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1208) q = kvzalloc(sizeof(*q), GFP_KERNEL | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1209) if (!q) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1210) *fg = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1211) goto fg_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1212) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1213) INIT_WORK(&q->u.work, z_erofs_decompressqueue_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1214) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1215) fg_out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1216) q = fgq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1217) init_waitqueue_head(&fgq->u.wait);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1218) atomic_set(&fgq->pending_bios, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1219) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1220) q->sb = sb;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1221) q->head = Z_EROFS_PCLUSTER_TAIL_CLOSED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1222) return q;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1223) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1224)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1225) /* define decompression jobqueue types */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1226) enum {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1227) JQ_BYPASS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1228) JQ_SUBMIT,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1229) NR_JOBQUEUES,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1230) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1231)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1232) static void *jobqueueset_init(struct super_block *sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1233) struct z_erofs_decompressqueue *q[],
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1234) struct z_erofs_decompressqueue *fgq, bool *fg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1235) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1236) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1237) * if managed cache is enabled, bypass jobqueue is needed,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1238) * no need to read from device for all pclusters in this queue.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1239) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1240) q[JQ_BYPASS] = jobqueue_init(sb, fgq + JQ_BYPASS, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1241) q[JQ_SUBMIT] = jobqueue_init(sb, fgq + JQ_SUBMIT, fg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1242)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1243) return tagptr_cast_ptr(tagptr_fold(tagptr1_t, q[JQ_SUBMIT], *fg));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1244) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1245)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1246) static void move_to_bypass_jobqueue(struct z_erofs_pcluster *pcl,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1247) z_erofs_next_pcluster_t qtail[],
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1248) z_erofs_next_pcluster_t owned_head)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1249) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1250) z_erofs_next_pcluster_t *const submit_qtail = qtail[JQ_SUBMIT];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1251) z_erofs_next_pcluster_t *const bypass_qtail = qtail[JQ_BYPASS];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1252)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1253) DBG_BUGON(owned_head == Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1254) if (owned_head == Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1255) owned_head = Z_EROFS_PCLUSTER_TAIL_CLOSED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1256)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1257) WRITE_ONCE(pcl->next, Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1258)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1259) WRITE_ONCE(*submit_qtail, owned_head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1260) WRITE_ONCE(*bypass_qtail, &pcl->next);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1261)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1262) qtail[JQ_BYPASS] = &pcl->next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1263) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1264)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1265) static void z_erofs_submit_queue(struct super_block *sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1266) struct z_erofs_decompress_frontend *f,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1267) struct list_head *pagepool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1268) struct z_erofs_decompressqueue *fgq,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1269) bool *force_fg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1270) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1271) struct erofs_sb_info *const sbi = EROFS_SB(sb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1272) z_erofs_next_pcluster_t qtail[NR_JOBQUEUES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1273) struct z_erofs_decompressqueue *q[NR_JOBQUEUES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1274) void *bi_private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1275) z_erofs_next_pcluster_t owned_head = f->clt.owned_head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1276) /* since bio will be NULL, no need to initialize last_index */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1277) pgoff_t last_index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1278) unsigned int nr_bios = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1279) struct bio *bio = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1280)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1281) bi_private = jobqueueset_init(sb, q, fgq, force_fg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1282) qtail[JQ_BYPASS] = &q[JQ_BYPASS]->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1283) qtail[JQ_SUBMIT] = &q[JQ_SUBMIT]->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1284)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1285) /* by default, all need io submission */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1286) q[JQ_SUBMIT]->head = owned_head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1287)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1288) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1289) struct z_erofs_pcluster *pcl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1290) pgoff_t cur, end;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1291) unsigned int i = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1292) bool bypass = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1293)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1294) /* no possible 'owned_head' equals the following */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1295) DBG_BUGON(owned_head == Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1296) DBG_BUGON(owned_head == Z_EROFS_PCLUSTER_NIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1297)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1298) pcl = container_of(owned_head, struct z_erofs_pcluster, next);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1299)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1300) cur = pcl->obj.index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1301) end = cur + pcl->pclusterpages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1302)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1303) /* close the main owned chain at first */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1304) owned_head = cmpxchg(&pcl->next, Z_EROFS_PCLUSTER_TAIL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1305) Z_EROFS_PCLUSTER_TAIL_CLOSED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1306)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1307) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1308) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1309)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1310) page = pickup_page_for_submission(pcl, i++, pagepool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1311) MNGD_MAPPING(sbi),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1312) GFP_NOFS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1313) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1314) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1315)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1316) if (bio && cur != last_index + 1) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1317) submit_bio_retry:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1318) submit_bio(bio);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1319) bio = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1320) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1321)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1322) if (!bio) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1323) bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1324)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1325) bio->bi_end_io = z_erofs_decompressqueue_endio;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1326) bio_set_dev(bio, sb->s_bdev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1327) bio->bi_iter.bi_sector = (sector_t)cur <<
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1328) LOG_SECTORS_PER_BLOCK;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1329) bio->bi_private = bi_private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1330) bio->bi_opf = REQ_OP_READ;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1331) if (f->readahead)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1332) bio->bi_opf |= REQ_RAHEAD;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1333) ++nr_bios;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1334) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1335)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1336) if (bio_add_page(bio, page, PAGE_SIZE, 0) < PAGE_SIZE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1337) goto submit_bio_retry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1338)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1339) last_index = cur;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1340) bypass = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1341) } while (++cur < end);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1342)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1343) if (!bypass)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1344) qtail[JQ_SUBMIT] = &pcl->next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1345) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1346) move_to_bypass_jobqueue(pcl, qtail, owned_head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1347) } while (owned_head != Z_EROFS_PCLUSTER_TAIL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1348)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1349) if (bio)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1350) submit_bio(bio);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1351)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1352) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1353) * although background is preferred, no one is pending for submission.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1354) * don't issue workqueue for decompression but drop it directly instead.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1355) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1356) if (!*force_fg && !nr_bios) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1357) kvfree(q[JQ_SUBMIT]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1358) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1359) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1360) z_erofs_decompress_kickoff(q[JQ_SUBMIT], *force_fg, nr_bios);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1361) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1362)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1363) static void z_erofs_runqueue(struct super_block *sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1364) struct z_erofs_decompress_frontend *f,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1365) struct list_head *pagepool, bool force_fg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1366) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1367) struct z_erofs_decompressqueue io[NR_JOBQUEUES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1368)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1369) if (f->clt.owned_head == Z_EROFS_PCLUSTER_TAIL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1370) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1371) z_erofs_submit_queue(sb, f, pagepool, io, &force_fg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1372)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1373) /* handle bypass queue (no i/o pclusters) immediately */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1374) z_erofs_decompress_queue(&io[JQ_BYPASS], pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1375)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1376) if (!force_fg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1377) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1378)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1379) /* wait until all bios are completed */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1380) io_wait_event(io[JQ_SUBMIT].u.wait,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1381) !atomic_read(&io[JQ_SUBMIT].pending_bios));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1382)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1383) /* handle synchronous decompress queue in the caller context */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1384) z_erofs_decompress_queue(&io[JQ_SUBMIT], pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1385) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1386)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1387) static int z_erofs_readpage(struct file *file, struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1388) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1389) struct inode *const inode = page->mapping->host;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1390) struct z_erofs_decompress_frontend f = DECOMPRESS_FRONTEND_INIT(inode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1391) int err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1392) LIST_HEAD(pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1393)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1394) trace_erofs_readpage(page, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1395)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1396) f.headoffset = (erofs_off_t)page->index << PAGE_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1397)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1398) err = z_erofs_do_read_page(&f, page, &pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1399) (void)z_erofs_collector_end(&f.clt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1400)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1401) /* if some compressed cluster ready, need submit them anyway */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1402) z_erofs_runqueue(inode->i_sb, &f, &pagepool, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1403)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1404) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1405) erofs_err(inode->i_sb, "failed to read, err [%d]", err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1406)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1407) if (f.map.mpage)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1408) put_page(f.map.mpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1409)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1410) /* clean up the remaining free pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1411) put_pages_list(&pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1412) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1413) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1414)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1415) static void z_erofs_readahead(struct readahead_control *rac)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1416) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1417) struct inode *const inode = rac->mapping->host;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1418) struct erofs_sb_info *const sbi = EROFS_I_SB(inode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1419)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1420) unsigned int nr_pages = readahead_count(rac);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1421) bool sync = (sbi->ctx.readahead_sync_decompress &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1422) nr_pages <= sbi->ctx.max_sync_decompress_pages);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1423) struct z_erofs_decompress_frontend f = DECOMPRESS_FRONTEND_INIT(inode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1424) struct page *page, *head = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1425) LIST_HEAD(pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1426)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1427) trace_erofs_readpages(inode, readahead_index(rac), nr_pages, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1428)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1429) f.readahead = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1430) f.headoffset = readahead_pos(rac);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1431)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1432) while ((page = readahead_page(rac))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1433) prefetchw(&page->flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1434)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1435) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1436) * A pure asynchronous readahead is indicated if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1437) * a PG_readahead marked page is hitted at first.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1438) * Let's also do asynchronous decompression for this case.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1439) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1440) sync &= !(PageReadahead(page) && !head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1441)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1442) set_page_private(page, (unsigned long)head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1443) head = page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1444) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1445)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1446) while (head) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1447) struct page *page = head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1448) int err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1449)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1450) /* traversal in reverse order */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1451) head = (void *)page_private(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1452)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1453) err = z_erofs_do_read_page(&f, page, &pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1454) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1455) erofs_err(inode->i_sb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1456) "readahead error at page %lu @ nid %llu",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1457) page->index, EROFS_I(inode)->nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1458) put_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1459) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1460)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1461) (void)z_erofs_collector_end(&f.clt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1462)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1463) z_erofs_runqueue(inode->i_sb, &f, &pagepool, sync);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1464)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1465) if (f.map.mpage)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1466) put_page(f.map.mpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1467)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1468) /* clean up the remaining free pages */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1469) put_pages_list(&pagepool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1470) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1471)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1472) const struct address_space_operations z_erofs_aops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1473) .readpage = z_erofs_readpage,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1474) .readahead = z_erofs_readahead,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1475) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1476)
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags