changeset 51601:de5bf3fe0233

revset: stop serializing node when using "%ln" Turning hundred of thousand of node from node to hex and back can be slow… what about we stop doing it? In many case were we are using node id we should be using revision id. However this is not a good reason to have a stupidly slow implementation of "%ln". This caught my attention again because the phase discovery during push make an extensive use of "%ln" or huge set. In absolute, that phase discovery probably should use "%ld" and need to improves its algorithmic complexity, but improving "%ln" seems simple and long overdue. This greatly speeds up `hg push` on repository with many drafts. Here are some relevant poulpe benchmarks: ### data-env-vars.name = mozilla-try-2023-03-22-zstd-sparse-revlog # benchmark.name = hg.command.push # bin-env-vars.hg.flavor = default # bin-env-vars.hg.py-re2-module = default # benchmark.variants.explicit-rev = all-out-heads # benchmark.variants.issue6528 = disabled # benchmark.variants.protocol = ssh # benchmark.variants.reuse-external-delta-parent = default ## benchmark.variants.revs = any-1-extra-rev before: 44.235070 after: 20.416329 (-53.85%, -23.82) ## benchmark.variants.revs = any-100-extra-rev before: 49.234697 after: 26.519829 (-46.14%, -22.71) ### benchmark.name = hg.command.bundle # bin-env-vars.hg.flavor = default # bin-env-vars.hg.py-re2-module = default # benchmark.variants.revs = all # benchmark.variants.type = none-streamv2 ## data-env-vars.name = heptapod-public-2024-03-25-zstd-sparse-revlog before: 10.138396 after: 7.750458 (-23.55%, -2.39) ## data-env-vars.name = mercurial-public-2024-03-22-zstd-sparse-revlog before: 1.263859 after: 0.700229 (-44.60%, -0.56) ## data-env-vars.name = mozilla-try-2023-03-22-zstd-sparse-revlog before: 399.484481 after: 346.5089 (-13.26%, -52.98) ## data-env-vars.name = pypy-2024-03-22-zstd-sparse-revlog before: 4.540080 after: 3.401700 (-25.07%, -1.14) ## data-env-vars.name = tryton-public-2024-03-22-zstd-sparse-revlog before: 2.975765 after: 1.870798 (-37.13%, -1.10)
author Pierre-Yves David <pierre-yves.david@octobus.net>
date Fri, 05 Apr 2024 11:05:54 +0200
parents f610d705c3ca
children b5d494f7d28a
files mercurial/revset.py mercurial/revsetlang.py
diffstat 2 files changed, 33 insertions(+), 3 deletions(-) [+]
line wrap: on
line diff
--- a/mercurial/revset.py	Tue Apr 09 14:41:48 2024 +0200
+++ b/mercurial/revset.py	Fri Apr 05 11:05:54 2024 +0200
@@ -149,6 +149,22 @@
         return x & subset
 
 
+def raw_node_set(repo, subset, x, order):
+    """argument is a list of nodeid, resolve and use them"""
+    nodes = _ordered_node_set(repo, x)
+    if order == followorder:
+        return subset & nodes
+    else:
+        return nodes & subset
+
+
+def _ordered_node_set(repo, nodes):
+    if not nodes:
+        return baseset()
+    to_rev = repo.changelog.index.rev
+    return baseset([to_rev(r) for r in nodes])
+
+
 def rangeset(repo, subset, x, y, order):
     m = getset(repo, fullreposet(repo), x)
     n = getset(repo, fullreposet(repo), y)
@@ -2772,6 +2788,7 @@
     b"parent": parentspec,
     b"parentpost": parentpost,
     b"smartset": rawsmartset,
+    b"nodeset": raw_node_set,
 }
 
 relations = {
--- a/mercurial/revsetlang.py	Tue Apr 09 14:41:48 2024 +0200
+++ b/mercurial/revsetlang.py	Fri Apr 05 11:05:54 2024 +0200
@@ -392,7 +392,7 @@
     elif op == b'negate':
         s = getstring(x[1], _(b"can't negate that"))
         return _analyze((b'string', b'-' + s))
-    elif op in (b'string', b'symbol', b'smartset'):
+    elif op in (b'string', b'symbol', b'smartset', b'nodeset'):
         return x
     elif op == b'rangeall':
         return (op, None)
@@ -441,8 +441,9 @@
         return 0, x
 
     op = x[0]
-    if op in (b'string', b'symbol', b'smartset'):
-        return 0.5, x  # single revisions are small
+    if op in (b'string', b'symbol', b'smartset', b'nodeset'):
+        # single revisions are small, and set of already computed revision are assumed to be cheap.
+        return 0.5, x
     elif op == b'and':
         wa, ta = _optimize(x[1])
         wb, tb = _optimize(x[2])
@@ -784,6 +785,8 @@
             if isinstance(arg, set):
                 arg = sorted(arg)
             ret.append(_formatintlist(list(arg)))
+        elif t == b'nodeset':
+            ret.append(_formatlistexp(list(arg), b"n"))
         else:
             raise error.ProgrammingError(b"unknown revspec item type: %r" % t)
     return b''.join(ret)
@@ -801,6 +804,10 @@
             newtree = (b'smartset', smartset.baseset(arg))
             inputs.append(newtree)
             ret.append(b"$")
+        elif t == b'nodeset':
+            newtree = (b'nodeset', arg)
+            inputs.append(newtree)
+            ret.append(b"$")
         else:
             raise error.ProgrammingError(b"unknown revspec item type: %r" % t)
     expr = b''.join(ret)
@@ -863,6 +870,12 @@
                 ret.append((b'baseset', arg))
                 pos += 1
                 continue
+            elif islist and d == b'n' and arg:
+                # we cannot turn the node into revision yet, but not
+                # serializing them will same a lot of time for large set.
+                ret.append((b'nodeset', arg))
+                pos += 1
+                continue
             try:
                 ret.append((None, f(list(arg), d)))
             except (TypeError, ValueError):