From 799e7cfc81e84a241bedd0ac812c6672bbd2e9e5 Mon Sep 17 00:00:00 2001
From: Christopher Haster <geky@geky.net>
Date: Mon, 19 Jun 2023 13:17:52 -0500
Subject: [PATCH] Flipped btree encoding to allow variable redund blocks

This should have been done as a part of the earlier tag reencoding work,
since having the block at the end was what allowed us to move the
redund-count out of the tag encoding.

New encoding:

  [-- 32-bit csum   --]
  [-- leb128 weight --]
  [-- leb128 trunk  --]
  [-- leb128 block  --]

Note that since our tags have an explicit size, we can store a variable
number of blocks. The plan is to use this to eventually store redundant
copies for error correction:

  [-- 32-bit csum   --]
  [-- leb128 weight --]
  [-- leb128 trunk  --]
  [-- leb128 block  --] -.
  [-- leb128 block  --]  +- n redundant blocks
  [-- leb128 block  --]  |
           ...          -'

This does have a significant tradeoff, we need to know the checksum size
to access the btree structure. This doesn't seem like a big deal, but
with the possibility of different checksum types may be an annoying
issue.

Note that FCRC was also flipped for consistency.
---
 lfs.c               | 42 ++++++++++++++++++++++--------------------
 scripts/dbgbtree.py |  8 ++++----
 scripts/dbgmtree.py |  8 ++++----
 3 files changed, 30 insertions(+), 28 deletions(-)

diff --git a/lfs.c b/lfs.c
index 4d698d0b..19f0f9e6 100644
--- a/lfs.c
+++ b/lfs.c
@@ -1338,8 +1338,8 @@ typedef struct lfsr_attr {
 
 // fcrc on-disk encoding
 typedef struct lfsr_fcrc {
-    lfs_size_t size;
     uint32_t crc;
+    lfs_size_t size;
 } lfsr_fcrc_t;
 
 // 1 leb128 + 1 crc32c => 9 bytes (worst case)
@@ -1349,28 +1349,29 @@ static lfs_ssize_t lfsr_fcrc_todisk(lfs_t *lfs, const lfsr_fcrc_t *fcrc,
         uint8_t buffer[static LFSR_FCRC_DSIZE]) {
     (void)lfs;
     lfs_ssize_t d = 0;
-    lfs_ssize_t d_ = lfs_toleb128(fcrc->size, &buffer[0], 5);
+
+    lfs_tole32_(fcrc->crc, &buffer[d]);
+    d += 4;
+
+    lfs_ssize_t d_ = lfs_toleb128(fcrc->size, &buffer[d], 5);
     if (d_ < 0) {
         return d_;
     }
     d += d_;
 
-    lfs_tole32_(fcrc->crc, &buffer[d]);
-    d += 4;
-
     return d;
 }
 
 static lfs_ssize_t lfsr_fcrc_fromdisk(lfs_t *lfs, lfsr_fcrc_t *fcrc,
         lfsr_data_t data) {
     lfs_ssize_t d = 0;
-    lfs_ssize_t d_ = lfsr_data_readleb128(lfs, data, d, &fcrc->size);
+    lfs_ssize_t d_ = lfsr_data_readle32(lfs, data, d, &fcrc->crc);
     if (d_ < 0) {
         return d_;
     }
     d += d_;
 
-    d_ = lfsr_data_readle32(lfs, data, d, &fcrc->crc);
+    d_ = lfsr_data_readleb128(lfs, data, d, &fcrc->size);
     if (d_ < 0) {
         return d_;
     }
@@ -2650,7 +2651,7 @@ static int lfsr_rbyd_commit(lfs_t *lfs, lfsr_rbyd_t *rbyd,
 
         // find the expected fcrc, don't bother avoiding a reread of the
         // perturb byte, as it should still be in our cache
-        lfsr_fcrc_t fcrc = {.size=lfs->cfg->prog_size, .crc=0};
+        lfsr_fcrc_t fcrc = {.crc=0, .size=lfs->cfg->prog_size};
         err = lfsr_bd_csum(lfs, rbyd_.block, aligned, lfs->cfg->prog_size,
                 lfs->cfg->prog_size,
                 &fcrc.crc);
@@ -3042,13 +3043,17 @@ static inline lfs_size_t lfsr_btree_setinlined(lfs_size_t weight) {
 
 // branch on-disk encoding
 
-// 3 leb128 + 1 crc32c => 19 bytes (worst case)
-#define LFSR_BRANCH_DSIZE (5+5+5+4)
+// 1 crc32c + 3 leb128 => 19 bytes (worst case)
+#define LFSR_BRANCH_DSIZE (4+5+5+5)
 
 static lfs_ssize_t lfsr_branch_todisk(lfs_t *lfs, const lfsr_rbyd_t *branch,
         uint8_t buffer[static LFSR_BRANCH_DSIZE]) {
     (void)lfs;
     lfs_ssize_t d = 0;
+
+    lfs_tole32_(branch->crc, &buffer[d]);
+    d += 4;
+
     lfs_ssize_t d_ = lfs_toleb128(branch->weight, &buffer[d], 5);
     if (d_ < 0) {
         return d_;
@@ -3067,9 +3072,6 @@ static lfs_ssize_t lfsr_branch_todisk(lfs_t *lfs, const lfsr_rbyd_t *branch,
     }
     d += d_;
 
-    lfs_tole32_(branch->crc, &buffer[d]);
-    d += 4;
-
     return d;
 }
 
@@ -3080,7 +3082,13 @@ static lfs_ssize_t lfsr_branch_fromdisk(lfs_t *lfs, lfsr_rbyd_t *branch,
     branch->off = 0;
 
     lfs_ssize_t d = 0;
-    lfs_ssize_t d_ = lfsr_data_readleb128(lfs, data, d, &branch->weight);
+    lfs_ssize_t d_ = lfsr_data_readle32(lfs, data, d, &branch->crc);
+    if (d_ < 0) {
+        return d_;
+    }
+    d += d_;
+
+    d_ = lfsr_data_readleb128(lfs, data, d, &branch->weight);
     if (d_ < 0) {
         return d_;
     }
@@ -3098,12 +3106,6 @@ static lfs_ssize_t lfsr_branch_fromdisk(lfs_t *lfs, lfsr_rbyd_t *branch,
     }
     d += d_;
 
-    d_ = lfsr_data_readle32(lfs, data, d, &branch->crc);
-    if (d_ < 0) {
-        return d_;
-    }
-    d += d_;
-
     return d;
 }
 
diff --git a/scripts/dbgbtree.py b/scripts/dbgbtree.py
index 501ae442..7eb54708 100755
--- a/scripts/dbgbtree.py
+++ b/scripts/dbgbtree.py
@@ -93,10 +93,10 @@ def fromtag(data):
     return tag>>15, tag&0x7fff, weight, size, 2+d+d_
 
 def frombtree(data):
-    w, d1 = fromleb128(data)
-    trunk, d2 = fromleb128(data[d1:])
-    block, d3 = fromleb128(data[d1+d2:])
-    crc = fromle32(data[d1+d2+d3:])
+    crc = fromle32(data)
+    w, d1 = fromleb128(data[4:])
+    trunk, d2 = fromleb128(data[4+d1:])
+    block, d3 = fromleb128(data[4+d1+d2:])
     return w, trunk, block, crc
 
 def popc(x):
diff --git a/scripts/dbgmtree.py b/scripts/dbgmtree.py
index a2f90b9e..9715c627 100755
--- a/scripts/dbgmtree.py
+++ b/scripts/dbgmtree.py
@@ -102,10 +102,10 @@ def frommdir(data):
     return blocks
 
 def frombtree(data):
-    w, d1 = fromleb128(data)
-    trunk, d2 = fromleb128(data[d1:])
-    block, d3 = fromleb128(data[d1+d2:])
-    crc = fromle32(data[d1+d2+d3:])
+    crc = fromle32(data)
+    w, d1 = fromleb128(data[4:])
+    trunk, d2 = fromleb128(data[4+d1:])
+    block, d3 = fromleb128(data[4+d1+d2:])
     return w, trunk, block, crc
 
 def popc(x):