PR#5899: user-friendly representation of backtraces#28
PR#5899: user-friendly representation of backtraces#28gasche wants to merge 9 commits intoocaml:trunkfrom
Conversation
stdlib/printexc.mli
Outdated
There was a problem hiding this comment.
This type is not documented, it should especially since it exposes a tuple rather than a record. Something Like
(** The type for backtrace slots.
{ul
{- [Known_location (is_raise, filename, line, schar, echar)] is for known
location where [is_raise] is [true] if ....}
{- [Unknown_location is_raise] is for unknown location where [is_raise]
is [true if] ... }}
|
I'm undecided about exposing backtrace_slot as a concrete type, since it will make it more difficult to extend the payload later with more information. What about exposing "getter" functions for each field instead? Or if we want it to be concrete (without waiting for inline records :-)), we could return it as a record type (using dummy values such as filename = "" when the location is unknown). |
|
I'm fine with getter functions. What about the following interface? |
|
What would "location" return when the location is unknown? It it returns a dummy value, we should probably do the same for lower-level functions, no? |
|
Also: the user might wonder about the purpose for Slot.t since getter functions could directly operate on row_slot. I understand the reason (avoiding to do the lookup for every call of a getter function). One possibility could be to hide this as an implementation detail by memoizing the last lookup, internally. But while this makes the API textually simpler, it also hides some performance properties that advanced users might prefer to now about. What do you think? |
|
Sorry, I meant the I think it's good to make people think about which kind of processing they intend to do on traces (bulk collection tasks or user information), and adapt the moment of conversion from raw to usable accordingly. @dbuenzli , as our API designer expert, do you have any opinion? |
|
Somehow I didn't get the messages from the further discussion. Regarding I would prefer Somehow I miss the point of having both a record and accessors. Besides accessors should not raise |
|
What's the point of the getter functions if the |
There are several changes:
- `raw_backtrace` is no longer an abstract type, but rather an
`raw_backtrace_slot array`, where `raw_backtrace_slot` is a new
abstract type. `raw_backtrace_slot` elements are hashable and
comparable. At runtime, values of this type contain either
a bytecode pointer or a frame_descr pointer. In order to prevent the
GC from walking through this pointer, the low-order bit is set to
1 when stored in the array.
- The old `loc_info` type is know public, renamed into `backtrace_slot`:
type backtrace_slot =
| Known_location of bool (* is_raise *)
* string (* filename *)
* int (* line number *)
* int (* start char *)
* int (* end char *)
| Unknown_location of bool (*is_raise*)
- new primitive :
val convert_raw_backtrace_slot: raw_backtrace_slot -> backtrace_slot
Rather than returning an option, it raises Failure when it is not
possible to get the debugging information. It seems more idiomatic,
especially because the exceptional case cannot appear only for a part
of the executable.
- the caml_convert_raw_backtrace primitive is removed; it is more
difficult to implement in the C side because of the new exception
interface described above.
- In the bytecode runtime, the events are no longer deserialized once
for each conversion, but once and for all at the first conversion,
and stored in a global array (*outside* the OCaml heap), sorted by
program counter value. I believe this information should not take
much memory in practice (it uses the same order of magnitude memory
as the bytecode executable). It also makes location lookup much more
efficient, as a dichomoty is used instead of linear search as
previously.
Jacques-Henri initially removed the primitive, which is deprecated since 4.01, but I suspect there still are uses in the wild. I guess we should wait for a few more versions.
Test the behavior of the deprecated primitive [caml_get_exception_backtrace], and minimal tests for hashing/comparison of raw backtrace slots.
- The internal [backtrace_slot] type is not exposed anymore, instead accessors function return orthogonal information (is_raise, location). This is both more extensible and more user-friendly. - The [raw_backtrace_slot] is exposed separately as a low-level type that most users should never use. The unsafety of marshalling is documented. Instead of defining [raw_backtrace = raw_backtrace_slot array], I kept [raw_backtrace] an abstract type with [length] and [get] functions for random-access. This should allow us to change the implementation in the future to be more robust wrt. marshalling (boxing the trace in a Custom block, or even possibly the raw slots at access time).
|
The most recent commit ( baa26a1 ) implements the discussed API, along with a few changes to take the remarks into account (no exceptions anymore). The type My plan is to merge this really soon (possibly tomorrow) so that 4.02 gets the runtime improvement (in particular the solution to PR#6302 that Alain needs). Luckily, the runtime side (the C code) hasn't been touched by the last commit, so it's a relatively low-risk change. |
Fix some tests
Correctly iter on static set of closures definitions
Cleanup external calls to caml_stw_empty_minor_heap
* Add support for "global" fields * Bootstrap
* Add support for "global" fields * Bootstrap
173842c Merge flambda-backend changes ed7eba2 Remove leading space from LINE. (oxcaml/oxcaml#484) bd61170 Bump magic numbers (ocaml#5) c50c47d Add CI builds with local allocations enabled 1412792 Move local allocations support behind '-extension local' 6d8e42a Better tail call behaviour in caml_applyN c7dac3d Typemod: toplevel bindings escape even if no variables are bound 82d6c3e Several fixes for partial application and currying d05c70c Pprintast support for new local syntax e0e62fc Typecheck x |> f y as (f y x), not ((f y) x) d7e34ce Remove autogeneration of @ocaml.curry b9a0593 Port oxcaml/oxcaml#493 0a872d9 Code review fixes from oxcaml/oxcaml#491 6c168bb Remove local allocation counting 3c6e7f0 Code review fixes from oxcaml/oxcaml#478 bb97207 Rename Lambda.apply_position a7cb650 Quieten Makefile when runtime dep files are not present c656dc9 Merge flambda-backend changes 11b5424 Avoid printing double spaces in function argument lists 7751faa Restore locations to Typedtree.{pat,let}_bound_idents_full e450b6c add build_ocaml_compiler.sexp 0403bb3 Revert PR 9895 to continue installing VERSION b3447db Ensure new local attributes are namespaced properly 7f213fc Allow empty functions again 8f22ad8 Bugfix: ensure local domain state is initialised 80f54dd Bugfix for Selectgen with regions e8133a1 Fix external-external signature inclusion 9840051 Bootstrap d879f23 Merge remote-tracking branch 'jane/local-reviewed' into local-merge 94454f5 Use Local_store for the local allocations ref 54a164c Create fewer regions, according to typechecking (ocaml#59) 1c2479b Merge flambda-backend changes ce34678 Fix printing of modes in return types 91f2281 Hook mode variable solving into Btype.snapshot/backtrack 54e4b09 Move Alloc_mode and Value_mode to Btype ff4611e Merge flambda-backend changes ce62e45 Ensure allocations are initialised, even dead ones 6b6ec5a Fix the alloc.ml test on 32-bit builds 81e9879 Merge flambda-backend changes 40a7f89 Update repo URL for ocaml-jst, and rename script. 0454ee7 Add some new locally-allocating primitives (ocaml#57) 8acdda1 Reset the local stack pointer in exception handlers (ocaml#56) 8dafa98 Improve typing for (||) and (&&) (ocaml#55) 8c64754 Fix make_check_all_arches (ocaml#54) b50cd45 Allow arguments to primitives to be local even in tail position (ocaml#53) cad125d Fix modes from or-patterns (ocaml#50) 4efdb72 Fix tailcalls tests with inlining (ocaml#52) 4a795cb Flambda support (ocaml#49) 74722cb Add [@ocaml.principal] and [@ocaml.noprincipal] attributes, and use in oo.mli 6d7d3b8 Ensure that functions are evaluated after their arguments (flambda-backend ocaml#353) 89bda6b Keep Sys.opaque_identity in Cmm and Mach (port upstream PR 9412) a39126a Fix tailcalls within regions (ocaml#48) 4ac4cfd Fix stdlib manpages build 3a95f5e Merge flambda-backend changes efe80c9 Add jane/pull-flambda-patches script fca94c4 Register allocations for Omitted parameter closures (ocaml#47) 103b139 Remove various FIXMEs (ocaml#46) 62ba2c1 Bootstrap a0062ad Allow local allocations for various primitives (ocaml#43) 7a2165e Allow primitives to be poly-moded (ocaml#43) 2af3f55 Fix a flaky test by refactoring TypePairs (ocaml#10638) 58dd807 Bootstrap ee3be10 Fix modes in build_apply for partial applications fe73656 Tweak for evaluation order of labelled partial applications (ocaml#10653) 0527570 Fix caml_modify on local allocations (ocaml#40) e657e99 Relax modes for `as` patterns (ocaml#42) f815bf2 Add special mode handling for tuples in matches and let bindings (ocaml#38) 39f1211 Only take the upper bounds of modes associated with allocations (ocaml#37) aec6fde Interpret arrow types in "local positions" differently c4f3319 Bootstrap ff6fdad Add some missing regions 40d586d Bootstrap 66d8110 Switch to a system with 3 modes for values f2c5a85 Bugfix for Comballoc with local allocations. (ocaml#41) 83bcd09 Fix bug with root scanning during compaction (ocaml#39) 1b5ec83 Track modes in Lambda.lfunction and onwards (ocaml#33) f1e2e97 Port ocaml#10728 56703cd Port ocaml#10081 eb66785 Support local allocations in i386 and fix amd64 bug (ocaml#31) c936b19 Disallow local recursive non-functions (ocaml#30) c7a193a GC support for local allocations (ocaml#29) 8dd7270 Nonlocal fields (ocaml#28) e19a2f0 Bootstrap 694b9ac Add syntax to the parser for local allocations (ocaml#26) f183008 Lower initial stack size 918226f Allow local closure allocations (ocaml#27) 2552e7d Introduce mode variables (ocaml#25) bc41c99 Minor fixes for local allocations (ocaml#24) a2a4e60 Runtime and compiler support for more local allocations (ocaml#23) d030554 Typechecking for local allocations (ocaml#21) 9ee2332 Bugfix missing from ocaml#20 02c4cef Retain block-structured local regions until Mach. 86dbe1c amd64: Move stack realloc calls out-of-line 324d218 More typing modes and locking of environments a4080b8 Initial version of local allocation (unsafe) git-subtree-dir: ocaml git-subtree-split: 173842c
0b0aefb Turn some partial application warnings into hints (ocaml#11338) (ocaml#30) 2caa9ee Add [@tail] and [@nontail] annotations on applications to control tailcalls (ocaml#31) 9fb218a Update `promote` target to use the `one` machinery (ocaml#28) b5ea912 Make empty types immediate bc08236 Add failing test of an empty type being immediate f2d439f Propagate escaping_context to Env locks to hint about errors (ocaml#25) 35569e1 Allow warning 68 to be controlled by attributes (ocaml#16) 28a6243 Allow type_argument to weaken return modes of expected function types (ocaml#24) cdc728f Fix 'make alldepend' in otherlibs/dynlink 7807d18 make alldepend 2d6af2f Merge flambda-backend changes git-subtree-dir: ocaml git-subtree-split: 0b0aefb
Add breadcrumb and header to docs pages
9 participants
Joint work with Jacques-Henri Jourdan, this is the latest patchset to improve backtrace collection and inspection. A previous version of this patch was submitted by Jacques-Henri on mantis (PR#5899); the new version has a few change suggestions, and we use the fancy github PR process in the hope of getting more eyeballs on the patch.
Main changes: