mu/045closure_name.cc

186 lines
7.2 KiB
C++
Raw Normal View History

//: Writing to a literal (not computed) address of 0 in a recipe chains two
//: spaces together. When a variable has a property of /space:1, it looks up
//: the variable in the chained/surrounding space. /space:2 looks up the
//: surrounding space of the surrounding space, etc.
4089 Clean up how we reclaim local scopes. It used to work like this (commit 3216): 1. Update refcounts of products after every instruction, EXCEPT: a) when instruction is a non-primitive and the callee starts with 'local-scope' (because it's already not decremented in 'return') OR: b) when instruction is primitive 'next-ingredient' or 'next-ingredient-without-typechecking', and its result is saved to a variable in the default space (because it's already incremented at the time of the call) 2. If a function starts with 'local-scope', force it to be reclaimed before each return. However, since locals may be returned, *very carefully* don't reclaim those. (See the logic in the old `escaping` and `should_update_refcount` functions.) However, this approach had issues. We needed two separate commands for 'local-scope' (reclaim locals on exit) and 'new-default-space' (programmer takes charge of reclaiming locals). The hard-coded reclamation duplicated refcounting logic. In addition to adding complexity, this implementation failed to work if a function overwrites default-space after setting up a local-scope (the old default-space is leaked). It also fails in the presence of continuations. Calling a continuation more than once was guaranteed to corrupt memory (commit 3986). After this commit, reclaiming local scopes now works like this: Update refcounts of products for every PRIMITIVE instruction. For non-primitive instructions, all the work happens in the `return` instruction: increment refcount of ingredients to `return` (unless -- one last bit of ugliness -- they aren't saved in the caller) decrement the refcount of the default-space use existing infrastructure for reclaiming as necessary if reclaiming default-space, first decrement refcount of each local again, use existing infrastructure for reclaiming as necessary This commit (finally!) completes the bulk[1] of step 2 of the plan in commit 3991. It was very hard until I gave up trying to tweak the existing implementation and just test-drove layer 43 from scratch. [1] There's still potential for memory corruption if we abuse `default-space`. I should probably try to add warnings about that at some point (todo in layer 45).
2017-10-23 06:14:19 +00:00
//:
2017-11-04 01:01:59 +00:00
//: todo: warn on default-space abuse. default-space for one recipe should
4089 Clean up how we reclaim local scopes. It used to work like this (commit 3216): 1. Update refcounts of products after every instruction, EXCEPT: a) when instruction is a non-primitive and the callee starts with 'local-scope' (because it's already not decremented in 'return') OR: b) when instruction is primitive 'next-ingredient' or 'next-ingredient-without-typechecking', and its result is saved to a variable in the default space (because it's already incremented at the time of the call) 2. If a function starts with 'local-scope', force it to be reclaimed before each return. However, since locals may be returned, *very carefully* don't reclaim those. (See the logic in the old `escaping` and `should_update_refcount` functions.) However, this approach had issues. We needed two separate commands for 'local-scope' (reclaim locals on exit) and 'new-default-space' (programmer takes charge of reclaiming locals). The hard-coded reclamation duplicated refcounting logic. In addition to adding complexity, this implementation failed to work if a function overwrites default-space after setting up a local-scope (the old default-space is leaked). It also fails in the presence of continuations. Calling a continuation more than once was guaranteed to corrupt memory (commit 3986). After this commit, reclaiming local scopes now works like this: Update refcounts of products for every PRIMITIVE instruction. For non-primitive instructions, all the work happens in the `return` instruction: increment refcount of ingredients to `return` (unless -- one last bit of ugliness -- they aren't saved in the caller) decrement the refcount of the default-space use existing infrastructure for reclaiming as necessary if reclaiming default-space, first decrement refcount of each local again, use existing infrastructure for reclaiming as necessary This commit (finally!) completes the bulk[1] of step 2 of the plan in commit 3991. It was very hard until I gave up trying to tweak the existing implementation and just test-drove layer 43 from scratch. [1] There's still potential for memory corruption if we abuse `default-space`. I should probably try to add warnings about that at some point (todo in layer 45).
2017-10-23 06:14:19 +00:00
//: never come from another, otherwise memory will be corrupted.
2015-04-17 18:22:59 +00:00
2015-04-24 17:19:03 +00:00
:(scenario closure)
def main [
2016-09-17 21:53:00 +00:00
default-space:space <- new location:type, 30
2:space/names:new-counter <- new-counter
10:num/raw <- increment-counter 2:space/names:new-counter
11:num/raw <- increment-counter 2:space/names:new-counter
]
def new-counter [
2016-09-17 21:53:00 +00:00
default-space:space <- new location:type, 30
x:num <- copy 23
y:num <- copy 13 # variable that will be incremented
2016-09-17 21:53:00 +00:00
return default-space:space
]
def increment-counter [
2016-09-17 21:53:00 +00:00
default-space:space <- new location:type, 30
0:space/names:new-counter <- next-ingredient # outer space must be created by 'new-counter' above
y:num/space:1 <- add y:num/space:1, 1 # increment
y:num <- copy 234 # dummy
return y:num/space:1
]
2015-10-29 19:04:30 +00:00
+name: lexically surrounding space for recipe increment-counter comes from new-counter
+mem: storing 15 in location 11
//: To make this work, compute the recipe that provides names for the
//: surrounding space of each recipe.
2015-04-17 18:22:59 +00:00
:(before "End Globals")
map<recipe_ordinal, recipe_ordinal> Surrounding_space; // internal to transform; no need to snapshot
:(before "End Reset")
Surrounding_space.clear();
:(before "Begin Type Modifying Transforms")
Transform.push_back(collect_surrounding_spaces); // idempotent
:(code)
void collect_surrounding_spaces(const recipe_ordinal r) {
trace(9991, "transform") << "--- collect surrounding spaces for recipe " << get(Recipe, r).name << end();
2016-10-20 05:10:35 +00:00
for (int i = 0; i < SIZE(get(Recipe, r).steps); ++i) {
const instruction& inst = get(Recipe, r).steps.at(i);
if (inst.is_label) continue;
2016-10-20 05:10:35 +00:00
for (int j = 0; j < SIZE(inst.products); ++j) {
2015-06-17 19:39:09 +00:00
if (is_literal(inst.products.at(j))) continue;
if (inst.products.at(j).name != "0") continue;
if (!is_mu_space(inst.products.at(j))) {
2016-09-17 21:43:13 +00:00
raise << "slot 0 should always have type address:array:location, but is '" << to_string(inst.products.at(j)) << "'\n" << end();
continue;
}
string_tree* s = property(inst.products.at(j), "names");
if (!s) {
raise << "slot 0 requires a /names property in recipe '" << get(Recipe, r).name << "'\n" << end();
continue;
}
3309 Rip out everything to fix one failing unit test (commit 3290; type abbreviations). This commit does several things at once that I couldn't come up with a clean way to unpack: A. It moves to a new representation for type trees without changing the actual definition of the `type_tree` struct. B. It adds unit tests for our type metadata precomputation, so that errors there show up early and in a simpler setting rather than dying when we try to load Mu code. C. It fixes a bug, guarding against infinite loops when precomputing metadata for recursive shape-shifting containers. To do this it uses a dumb way of comparing type_trees, comparing their string representations instead. That is likely incredibly inefficient. Perhaps due to C, this commit has made Mu incredibly slow. Running all tests for the core and the edit/ app now takes 6.5 minutes rather than 3.5 minutes. == more notes and details I've been struggling for the past week now to back out of a bad design decision, a premature optimization from the early days: storing atoms directly in the 'value' slot of a cons cell rather than creating a special 'atom' cons cell and storing it on the 'left' slot. In other words, if a cons cell looks like this: o / | \ left val right ..then the type_tree (a b c) used to look like this (before this commit): o | \ a o | \ b o | \ c null ..rather than like this 'classic' approach to s-expressions which never mixes val and right (which is what we now have): o / \ o o | / \ a o o | / \ b o null | c The old approach made several operations more complicated, most recently the act of replacing a (possibly atom/leaf) sub-tree with another. That was the final straw that got me to realize the contortions I was going through to save a few type_tree nodes (cons cells). Switching to the new approach was hard partly because I've been using the old approach for so long and type_tree manipulations had pervaded everything. Another issue I ran into was the realization that my layers were not cleanly separated. Key parts of early layers (precomputing type metadata) existed purely for far later ones (shape-shifting types). Layers I got repeatedly stuck at: 1. the transform for precomputing type sizes (layer 30) 2. type-checks on merge instructions (layer 31) 3. the transform for precomputing address offsets in types (layer 36) 4. replace operations in supporting shape-shifting recipes (layer 55) After much thrashing I finally noticed that it wasn't the entirety of these layers that was giving me trouble, but just the type metadata precomputation, which had bugs that weren't manifesting until 30 layers later. Or, worse, when loading .mu files before any tests had had a chance to run. A common failure mode was running into types at run time that I hadn't precomputed metadata for at transform time. Digging into these bugs got me to realize that what I had before wasn't really very good, but a half-assed heuristic approach that did a whole lot of extra work precomputing metadata for utterly meaningless types like `((address number) 3)` which just happened to be part of a larger type like `(array (address number) 3)`. So, I redid it all. I switched the representation of types (because the old representation made unit tests difficult to retrofit) and added unit tests to the metadata precomputation. I also made layer 30 only do the minimal metadata precomputation it needs for the concepts introduced until then. In the process, I also made the precomputation more correct than before, and added hooks in the right place so that I could augment the logic when I introduced shape-shifting containers. == lessons learned There's several levels of hygiene when it comes to layers: 1. Every layer introduces precisely what it needs and in the simplest way possible. If I was building an app until just that layer, nothing would seem over-engineered. 2. Some layers are fore-shadowing features in future layers. Sometimes this is ok. For example, layer 10 foreshadows containers and arrays and so on without actually supporting them. That is a net win because it lets me lay out the core of Mu's data structures out in one place. But if the fore-shadowing gets too complex things get nasty. Not least because it can be hard to write unit tests for features before you provide the plumbing to visualize and manipulate them. 3. A layer is introducing features that are tested only in later layers. 4. A layer is introducing features with tests that are invalidated in later layers. (This I knew from early on to be an obviously horrendous idea.) Summary: avoid Level 2 (foreshadowing layers) as much as possible. Tolerate it indefinitely for small things where the code stays simple over time, but become strict again when things start to get more complex. Level 3 is mostly a net lose, but sometimes it can be expedient (a real case of the usually grossly over-applied term "technical debt"), and it's better than the conventional baseline of no layers and no scenarios. Just clean it up as soon as possible. Definitely avoid layer 4 at any time. == minor lessons Avoid unit tests for trivial things, write scenarios in context as much as possible. But within those margins unit tests are fine. Just introduce them before any scenarios (commit 3297). Reorganizing layers can be easy. Just merge layers for starters! Punt on resplitting them in some new way until you've gotten them to work. This is the wisdom of Refactoring: small steps. What made it hard was not wanting to merge *everything* between layer 30 and 55. The eventual insight was realizing I just need to move those two full-strength transforms and nothing else.
2016-09-10 01:32:52 +00:00
if (!s->atom) raise << "slot 0 should have a single value in /names, but got '" << to_string(inst.products.at(j)) << "'\n" << end();
const string& surrounding_recipe_name = s->value;
2016-02-06 17:29:36 +00:00
if (surrounding_recipe_name.empty()) {
raise << "slot 0 doesn't initialize its /names property in recipe '" << get(Recipe, r).name << "'\n" << end();
2016-02-06 17:29:36 +00:00
continue;
}
if (contains_key(Surrounding_space, r)
&& get(Surrounding_space, r) != get(Recipe_ordinal, surrounding_recipe_name)) {
raise << "recipe '" << get(Recipe, r).name << "' can have only one 'surrounding' recipe but has '" << get(Recipe, get(Surrounding_space, r)).name << "' and '" << surrounding_recipe_name << "'\n" << end();
continue;
}
trace(9993, "name") << "lexically surrounding space for recipe " << get(Recipe, r).name << " comes from " << surrounding_recipe_name << end();
if (!contains_key(Recipe_ordinal, surrounding_recipe_name)) {
raise << "can't find recipe providing surrounding space for '" << get(Recipe, r).name << "'; looking for '" << surrounding_recipe_name << "'\n" << end();
continue;
}
put(Surrounding_space, r, get(Recipe_ordinal, surrounding_recipe_name));
}
}
}
//: Once surrounding spaces are available, transform_names uses them to handle
//: /space properties.
:(replace{} "int lookup_name(const reagent& r, const recipe_ordinal default_recipe)")
int lookup_name(const reagent& x, const recipe_ordinal default_recipe) {
if (!has_property(x, "space")) {
2016-02-26 21:04:55 +00:00
if (Name[default_recipe].empty()) raise << "name not found: " << x.name << '\n' << end();
return Name[default_recipe][x.name];
}
string_tree* p = property(x, "space");
3309 Rip out everything to fix one failing unit test (commit 3290; type abbreviations). This commit does several things at once that I couldn't come up with a clean way to unpack: A. It moves to a new representation for type trees without changing the actual definition of the `type_tree` struct. B. It adds unit tests for our type metadata precomputation, so that errors there show up early and in a simpler setting rather than dying when we try to load Mu code. C. It fixes a bug, guarding against infinite loops when precomputing metadata for recursive shape-shifting containers. To do this it uses a dumb way of comparing type_trees, comparing their string representations instead. That is likely incredibly inefficient. Perhaps due to C, this commit has made Mu incredibly slow. Running all tests for the core and the edit/ app now takes 6.5 minutes rather than 3.5 minutes. == more notes and details I've been struggling for the past week now to back out of a bad design decision, a premature optimization from the early days: storing atoms directly in the 'value' slot of a cons cell rather than creating a special 'atom' cons cell and storing it on the 'left' slot. In other words, if a cons cell looks like this: o / | \ left val right ..then the type_tree (a b c) used to look like this (before this commit): o | \ a o | \ b o | \ c null ..rather than like this 'classic' approach to s-expressions which never mixes val and right (which is what we now have): o / \ o o | / \ a o o | / \ b o null | c The old approach made several operations more complicated, most recently the act of replacing a (possibly atom/leaf) sub-tree with another. That was the final straw that got me to realize the contortions I was going through to save a few type_tree nodes (cons cells). Switching to the new approach was hard partly because I've been using the old approach for so long and type_tree manipulations had pervaded everything. Another issue I ran into was the realization that my layers were not cleanly separated. Key parts of early layers (precomputing type metadata) existed purely for far later ones (shape-shifting types). Layers I got repeatedly stuck at: 1. the transform for precomputing type sizes (layer 30) 2. type-checks on merge instructions (layer 31) 3. the transform for precomputing address offsets in types (layer 36) 4. replace operations in supporting shape-shifting recipes (layer 55) After much thrashing I finally noticed that it wasn't the entirety of these layers that was giving me trouble, but just the type metadata precomputation, which had bugs that weren't manifesting until 30 layers later. Or, worse, when loading .mu files before any tests had had a chance to run. A common failure mode was running into types at run time that I hadn't precomputed metadata for at transform time. Digging into these bugs got me to realize that what I had before wasn't really very good, but a half-assed heuristic approach that did a whole lot of extra work precomputing metadata for utterly meaningless types like `((address number) 3)` which just happened to be part of a larger type like `(array (address number) 3)`. So, I redid it all. I switched the representation of types (because the old representation made unit tests difficult to retrofit) and added unit tests to the metadata precomputation. I also made layer 30 only do the minimal metadata precomputation it needs for the concepts introduced until then. In the process, I also made the precomputation more correct than before, and added hooks in the right place so that I could augment the logic when I introduced shape-shifting containers. == lessons learned There's several levels of hygiene when it comes to layers: 1. Every layer introduces precisely what it needs and in the simplest way possible. If I was building an app until just that layer, nothing would seem over-engineered. 2. Some layers are fore-shadowing features in future layers. Sometimes this is ok. For example, layer 10 foreshadows containers and arrays and so on without actually supporting them. That is a net win because it lets me lay out the core of Mu's data structures out in one place. But if the fore-shadowing gets too complex things get nasty. Not least because it can be hard to write unit tests for features before you provide the plumbing to visualize and manipulate them. 3. A layer is introducing features that are tested only in later layers. 4. A layer is introducing features with tests that are invalidated in later layers. (This I knew from early on to be an obviously horrendous idea.) Summary: avoid Level 2 (foreshadowing layers) as much as possible. Tolerate it indefinitely for small things where the code stays simple over time, but become strict again when things start to get more complex. Level 3 is mostly a net lose, but sometimes it can be expedient (a real case of the usually grossly over-applied term "technical debt"), and it's better than the conventional baseline of no layers and no scenarios. Just clean it up as soon as possible. Definitely avoid layer 4 at any time. == minor lessons Avoid unit tests for trivial things, write scenarios in context as much as possible. But within those margins unit tests are fine. Just introduce them before any scenarios (commit 3297). Reorganizing layers can be easy. Just merge layers for starters! Punt on resplitting them in some new way until you've gotten them to work. This is the wisdom of Refactoring: small steps. What made it hard was not wanting to merge *everything* between layer 30 and 55. The eventual insight was realizing I just need to move those two full-strength transforms and nothing else.
2016-09-10 01:32:52 +00:00
if (!p || !p->atom) raise << "/space property should have exactly one (non-negative integer) value\n" << end();
int n = to_integer(p->value);
assert(n >= 0);
recipe_ordinal surrounding_recipe = lookup_surrounding_recipe(default_recipe, n);
if (surrounding_recipe == -1) return -1;
set<recipe_ordinal> done;
vector<recipe_ordinal> path;
return lookup_name(x, surrounding_recipe, done, path);
}
// If the recipe we need to lookup this name in doesn't have names done yet,
// recursively call transform_names on it.
int lookup_name(const reagent& x, const recipe_ordinal r, set<recipe_ordinal>& done, vector<recipe_ordinal>& path) {
if (!Name[r].empty()) return Name[r][x.name];
if (contains_key(done, r)) {
raise << "can't compute address of '" << to_string(x) << "' because\n" << end();
2016-10-20 05:10:35 +00:00
for (int i = 1; i < SIZE(path); ++i) {
2016-02-26 21:04:55 +00:00
raise << path.at(i-1) << " requires computing names of " << path.at(i) << '\n' << end();
}
2016-02-26 21:04:55 +00:00
raise << path.at(SIZE(path)-1) << " requires computing names of " << r << "..ad infinitum\n" << end();
return -1;
}
done.insert(r);
path.push_back(r);
transform_names(r); // Not passing 'done' through. Might this somehow cause an infinite loop?
assert(!Name[r].empty());
return Name[r][x.name];
}
recipe_ordinal lookup_surrounding_recipe(const recipe_ordinal r, int n) {
if (n == 0) return r;
if (!contains_key(Surrounding_space, r)) {
raise << "don't know surrounding recipe of '" << get(Recipe, r).name << "'\n" << end();
return -1;
}
assert(contains_key(Surrounding_space, r));
return lookup_surrounding_recipe(get(Surrounding_space, r), n-1);
}
//: weaken use-before-set detection just a tad
:(replace{} "bool already_transformed(const reagent& r, const map<string, int>& names)")
bool already_transformed(const reagent& r, const map<string, int>& names) {
if (has_property(r, "space")) {
string_tree* p = property(r, "space");
3309 Rip out everything to fix one failing unit test (commit 3290; type abbreviations). This commit does several things at once that I couldn't come up with a clean way to unpack: A. It moves to a new representation for type trees without changing the actual definition of the `type_tree` struct. B. It adds unit tests for our type metadata precomputation, so that errors there show up early and in a simpler setting rather than dying when we try to load Mu code. C. It fixes a bug, guarding against infinite loops when precomputing metadata for recursive shape-shifting containers. To do this it uses a dumb way of comparing type_trees, comparing their string representations instead. That is likely incredibly inefficient. Perhaps due to C, this commit has made Mu incredibly slow. Running all tests for the core and the edit/ app now takes 6.5 minutes rather than 3.5 minutes. == more notes and details I've been struggling for the past week now to back out of a bad design decision, a premature optimization from the early days: storing atoms directly in the 'value' slot of a cons cell rather than creating a special 'atom' cons cell and storing it on the 'left' slot. In other words, if a cons cell looks like this: o / | \ left val right ..then the type_tree (a b c) used to look like this (before this commit): o | \ a o | \ b o | \ c null ..rather than like this 'classic' approach to s-expressions which never mixes val and right (which is what we now have): o / \ o o | / \ a o o | / \ b o null | c The old approach made several operations more complicated, most recently the act of replacing a (possibly atom/leaf) sub-tree with another. That was the final straw that got me to realize the contortions I was going through to save a few type_tree nodes (cons cells). Switching to the new approach was hard partly because I've been using the old approach for so long and type_tree manipulations had pervaded everything. Another issue I ran into was the realization that my layers were not cleanly separated. Key parts of early layers (precomputing type metadata) existed purely for far later ones (shape-shifting types). Layers I got repeatedly stuck at: 1. the transform for precomputing type sizes (layer 30) 2. type-checks on merge instructions (layer 31) 3. the transform for precomputing address offsets in types (layer 36) 4. replace operations in supporting shape-shifting recipes (layer 55) After much thrashing I finally noticed that it wasn't the entirety of these layers that was giving me trouble, but just the type metadata precomputation, which had bugs that weren't manifesting until 30 layers later. Or, worse, when loading .mu files before any tests had had a chance to run. A common failure mode was running into types at run time that I hadn't precomputed metadata for at transform time. Digging into these bugs got me to realize that what I had before wasn't really very good, but a half-assed heuristic approach that did a whole lot of extra work precomputing metadata for utterly meaningless types like `((address number) 3)` which just happened to be part of a larger type like `(array (address number) 3)`. So, I redid it all. I switched the representation of types (because the old representation made unit tests difficult to retrofit) and added unit tests to the metadata precomputation. I also made layer 30 only do the minimal metadata precomputation it needs for the concepts introduced until then. In the process, I also made the precomputation more correct than before, and added hooks in the right place so that I could augment the logic when I introduced shape-shifting containers. == lessons learned There's several levels of hygiene when it comes to layers: 1. Every layer introduces precisely what it needs and in the simplest way possible. If I was building an app until just that layer, nothing would seem over-engineered. 2. Some layers are fore-shadowing features in future layers. Sometimes this is ok. For example, layer 10 foreshadows containers and arrays and so on without actually supporting them. That is a net win because it lets me lay out the core of Mu's data structures out in one place. But if the fore-shadowing gets too complex things get nasty. Not least because it can be hard to write unit tests for features before you provide the plumbing to visualize and manipulate them. 3. A layer is introducing features that are tested only in later layers. 4. A layer is introducing features with tests that are invalidated in later layers. (This I knew from early on to be an obviously horrendous idea.) Summary: avoid Level 2 (foreshadowing layers) as much as possible. Tolerate it indefinitely for small things where the code stays simple over time, but become strict again when things start to get more complex. Level 3 is mostly a net lose, but sometimes it can be expedient (a real case of the usually grossly over-applied term "technical debt"), and it's better than the conventional baseline of no layers and no scenarios. Just clean it up as soon as possible. Definitely avoid layer 4 at any time. == minor lessons Avoid unit tests for trivial things, write scenarios in context as much as possible. But within those margins unit tests are fine. Just introduce them before any scenarios (commit 3297). Reorganizing layers can be easy. Just merge layers for starters! Punt on resplitting them in some new way until you've gotten them to work. This is the wisdom of Refactoring: small steps. What made it hard was not wanting to merge *everything* between layer 30 and 55. The eventual insight was realizing I just need to move those two full-strength transforms and nothing else.
2016-09-10 01:32:52 +00:00
if (!p || !p->atom) {
raise << "/space property should have exactly one (non-negative integer) value in '" << r.original_string << "'\n" << end();
return false;
}
if (p->value != "0") return true;
}
return contains_key(names, r.name);
}
:(scenario missing_surrounding_space)
% Hide_errors = true;
def f [
local-scope
x:num/space:1 <- copy 34
]
+error: don't know surrounding recipe of 'f'
+error: f: can't find a place to store 'x'
2016-05-19 00:05:22 +00:00
//: extra test for try_reclaim_locals() from previous layers
:(scenario local_scope_ignores_nonlocal_spaces)
def new-scope [
4089 Clean up how we reclaim local scopes. It used to work like this (commit 3216): 1. Update refcounts of products after every instruction, EXCEPT: a) when instruction is a non-primitive and the callee starts with 'local-scope' (because it's already not decremented in 'return') OR: b) when instruction is primitive 'next-ingredient' or 'next-ingredient-without-typechecking', and its result is saved to a variable in the default space (because it's already incremented at the time of the call) 2. If a function starts with 'local-scope', force it to be reclaimed before each return. However, since locals may be returned, *very carefully* don't reclaim those. (See the logic in the old `escaping` and `should_update_refcount` functions.) However, this approach had issues. We needed two separate commands for 'local-scope' (reclaim locals on exit) and 'new-default-space' (programmer takes charge of reclaiming locals). The hard-coded reclamation duplicated refcounting logic. In addition to adding complexity, this implementation failed to work if a function overwrites default-space after setting up a local-scope (the old default-space is leaked). It also fails in the presence of continuations. Calling a continuation more than once was guaranteed to corrupt memory (commit 3986). After this commit, reclaiming local scopes now works like this: Update refcounts of products for every PRIMITIVE instruction. For non-primitive instructions, all the work happens in the `return` instruction: increment refcount of ingredients to `return` (unless -- one last bit of ugliness -- they aren't saved in the caller) decrement the refcount of the default-space use existing infrastructure for reclaiming as necessary if reclaiming default-space, first decrement refcount of each local again, use existing infrastructure for reclaiming as necessary This commit (finally!) completes the bulk[1] of step 2 of the plan in commit 3991. It was very hard until I gave up trying to tweak the existing implementation and just test-drove layer 43 from scratch. [1] There's still potential for memory corruption if we abuse `default-space`. I should probably try to add warnings about that at some point (todo in layer 45).
2017-10-23 06:14:19 +00:00
local-scope
2016-09-17 19:55:10 +00:00
x:&:num <- new number:type
*x:&:num <- copy 34
2016-09-17 21:53:00 +00:00
return default-space:space
2016-05-19 00:05:22 +00:00
]
def use-scope [
local-scope
outer:space/names:new-scope <- next-ingredient
2016-09-17 21:53:00 +00:00
0:space/names:new-scope <- copy outer:space
2016-09-17 19:55:10 +00:00
return *x:&:num/space:1
2016-05-19 00:05:22 +00:00
]
def main [
2016-09-17 21:53:00 +00:00
1:space/raw <- new-scope
3:num/raw <- use-scope 1:space/raw
2016-05-19 00:05:22 +00:00
]
+mem: storing 34 in location 3
2018-06-25 17:59:20 +00:00
:(scenario recursive_transform_names)
def foo [
local-scope
x:num <- copy 0
2018-06-25 20:54:48 +00:00
return default-space:space/names:foo
2018-06-25 17:59:20 +00:00
]
def main [
local-scope
0:space/names:foo <- foo
x:num/space:1 <- copy 34
]
$error: 0