Enable all emscripten file systems: IDBFS, NODEFS, PROXYFS and WORKERFS#1596
Enable all emscripten file systems: IDBFS, NODEFS, PROXYFS and WORKERFS#1596hoodmane merged 14 commits intopyodide:mainfrom
Conversation
|
With the ba6915e API in place, it works from the console:
FF doesn't expose |
|
Thanks for working on this @bollwyvl: very interesting! One comment: although persistence is definitely something that would be good in some use cases, in others the lack of persistence with MEMFS might be preferable. Or at least more work would be necessary to manage file system state otherwise. For instance, for teaching Python it's appealing to know that users can re-start the tab and restart from from a clean env if necessary. Same for some small demos where there is no need for persistence. So it might be good to support passing Another thing I'm curious about is performance. I would imagine that MEMFS would be faster since no actual disk operations need to happen, though it probably also depends on how efficient both implementations are. |
|
@rth I think the idea here is that you use MEMFS for everything except some particular directory which is mounted as a persistent directory. I believe that the model supported by Emscripten is: no matter what, the main file system is MEMFS. Inside of MEMFS it is possible to set up directories that sync to other file systems. |
|
Right, as this PR currently stands, it's an entirely opt-in API, and layers on top of the existing, required From a performance perspective: I don't think it's a big driver in the interactive state, as the sync in-and-out is not automagical in any way. A simple usage might choose to only pay the |
|
OK, then great! I missed your example. The size impact seem negligible.
Could you elaborate on the motivation there? If we can avoid re-exporting and documenting a significant part of Emscripten's File system API (and use it directly instead) it would be preferable IMO. |
|
Again, this is my first time out with the emscripten plumbing. I don't have
a solid feel for what would be available at run time to a downstream or
user if it's not declared within the thing being compiled, but only that
the various *FS have to be compiled in.
If there is a pyodide.Emscripten hanging out someplace I don't know about,
then wrapping these three of the tens of FS members is silly, but for
example, making syncfs into an awaitable makes it seem more like the rest
of the pyodide public API.
… |
|
We haven't exposed the emscripten filesystem as a part of our public API. There is desire downstream to be able to directly adjust the filesystem, so I think we should expose it. It would make sense to add |
Sure, that sounds good! I had a look at the typings for emscripten, and they seem widely-used, but pretty out-of-date, but perhaps the
Oh, and hiding down at the bottom of the API, i did find |
|
@bollwyvl Can this be marked as "Ready for review" so it can be merged? Thanks! |
|
I'll also work up a PR just exposing the as-documented |
|
Ar the fails actually indicating anything? Also, regarding the notional Detailsconfigure: ./configure --host=wasm32-unknown-linux --prefix=~/pyodide/cpython/build/libffi/target --enable-static --disable-shared --disable-dependency-tracking --disable-builddir --disable-multi-os-directory --disable-raw-api checking build system type... x86_64-pc-linux-gnu checking host system type... wasm32-unknown-linux-gnu checking target system type... wasm32-unknown-linux-gnu checking for gsed... sed checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for wasm32-unknown-linux-strip... no checking for strip... strip checking for a thread-safe mkdir -p... /usr/bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking whether make supports nested variables... yes checking for wasm32-unknown-linux-gcc... ~/pyodide/emsdk/emsdk/upstream/emscripten/emcc checking whether the C compiler works... no configure: error: in `~/pyodide/cpython/build/libffi': configure: error: C compiler cannot create executables See `config.log' for more details make[1]: *** [Makefile:120: ~/pyodide/cpython/installs/python-3.9.5/lib/libffi.a] Error 77 make[1]: Leaving directory '~/pyodide/cpython' make: *** [Makefile:205: cpython/installs/python-3.9.5/lib/python3.9] Error 2 where I don't know enough about the tea leaves to unstick, so would not hold my breath on it for me. On the up-side: the Emscripten typings were recently updated... |
|
Ah, more digging: |
|
The CI failures so far are just the normal ones. CircleCi is a bit flakey for us, the webworker tests in particular fail randomly all the time. Is there a reason you don't want to use the docker image to do the local install? It should prevent these issues. Though it's good for people to try to build natively occasionally so that we can update the instructions. |
|
What does |
|
haha, would not have found that. This is getting me... further... diff --git a/cpython/Makefile b/cpython/Makefile
index 4d0399a..2741c4e 100644
--- a/cpython/Makefile
+++ b/cpython/Makefile
@@ -117,7 +117,9 @@ $(INSTALL)/lib/libffi.a :
rm -rf $(FFIBUILD)
git clone $(LIBFFIREPO) $(FFIBUILD) --shallow-exclude upstream-base
source $(PYODIDE_ROOT)/emsdk/emsdk/emsdk_env.sh
- cd $(FFIBUILD) && git checkout $(LIBFFI_COMMIT) && ./build.sh --pyodide-fpcast && make install
+ cd $(FFIBUILD) && git checkout $(LIBFFI_COMMIT)
+ cd $(FFIBUILD) && sed -i 's/_main,_malloc,_free/["_main","_malloc","_free"]/' build.sh
+ cd $(FFIBUILD) && ./build.sh --pyodide-fpcast && make install
cp $(FFIBUILD)/target/include/*.h $(BUILD)/Include/
mkdir -p $(INSTALL)/lib
cp $(FFIBUILD)/target/lib/libffi.a $(INSTALL)/lib/
|
… but got `<class 'str'>` See discussion on pyodide/pyodide#1596
|
#1692 is up, with the strawman API name of |
Yeah, I have to deal with all kinds of container junk at
Welp, anytime PRs comes from me, they'll at least have sorta built locally, once! I generally try to make the |
|
Set it to FS. As a light indicator, the ipdfs test reads a bit more nicely. |
|
Down to just |
|
Yup CI is doing good enough to merge now if it passes review. You have only 4/6 checks marked, you want this reviewed? A convenience API could probably be added later, might be nice to have something in the docs about this stuff though. |
Sure, closed that out and opened a ticket. |
As mentioned... I haven't been able to play with the API much, but will... once there are nightlies with persistence. It'll be a lot easier for me to start rapping on examples once I've fought some of our (self-made) problems on jupyterlite. As it is, the API should have fairly good SEO with To that end, is there a "fast path" for docs-only PRs that wouldn't invoke the entire CI pipeline? |
|
Putting |
|
Right updating the docs later makes sense to me. Not much harm in having a feature that nobody knows about for a little while. I guess the fear is that if people get the features they want added in they may never actually come back to update the docs. |
don't i know it, friend! I added a note to that effect: #1715. Basically, it would be my preference if we could get some idea of what people might do with it (after doing some of it) the Hard Way, make a better API (especially for people actually using pyodide-flavored python) and land it here, rather than trying to do it our darned selves on |
|
So i guess i've argued myself out of the last two bullets, so this is probably ready for review! |
|
Okay I'll take a look when I have time. |
|
If you resolve the merge conflict I will merge this. |
|
I'm still having some issues with the node-based tests locally, but everything else appears fine, locally 😁. Here's hoping! |
|
ha, I'll take it: |
|
Thanks for working through this with me! Looking forward to exploring how
we can use it to do fun stuff!
|
|
Thanks for contributing! |
|
Guys I can just say that you are doing some magic here 😁 |
* added build script * Apply libffi-emscripten patch * Some changes to wasm32/ffi.c * Remove exit(0); from test suites * Fix LONGDOUBLE argument type * Use more macros in ffi.c * Use switch statements instead of if chains * Implemented struct args * Finish struct implementation * Partially working closures * Got closures working (most of closures test suite passes) * Revert changes to test suite * Update .gitignore * Apply code formatter * Use stackSave and stackRestore rather than directly adjusting stack pointer * Add missing break * Fix visibility of ffi_closure_alloc and ffi_closure_free * Fix FFI_TYPE_STRUCT and FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used sig needs to be vi here for FFI_TYPE_STRUCT and FFI_TYPE_LONGDOUBLE, noticed this while running the test suite without WASM_BIGINT support. * Always use dynCall rather than direct wasmTable lookup (function pointer cast emulation changes dynCall) * Prevent closures.c from duplicating symbols * Try to set up CI * Add test with bigint * Make test methods static * Remove BigInt shorthand because it messes up terser * Add selenium tests * Update tests a bit to try to make CI work * WASM_BIGINT is a linker flag not a compile flag * Finish getting CI working (#1) * update gitignore * Avoid adding "use strict;" to generated JS This should be controlled by -s STRICT_JS in Emscripten. * Make JavaScript ES5 compliant * Remove redundant EXPORTED_RUNTIME_METHODS settings * Fix definition of DEREF_I16 * Avoid marshalling FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used * Add missing FFI_TYPE_STRUCT signature * Improve test scripts * Remove redundant EXPORTED_RUNTIME_METHODS settings * Add missing EOL * Add struct unpacking tests * Update ci config to try to actually use WASM_BIGINT * Revert "Avoid marshalling FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used" This reverts commit 61bd5a3. * Fix single_entry_structs tests * Fix return from closure call * Fix 64 bit return from closures * only allocate as much space on stack for return pointer as needed * Revert "only allocate as much space on stack for return pointer as needed" This reverts commit e54a30f. * xfail two tests * Fix err_bad_abi test * Remove test logging junk * Try to set up long double marshalling for closures * xfail err_bad_abi * Fix reference errors in previous commit * Add missing argument pointer assignment * Fix signature of function pointer in cls_dbls_struct * Fix longdouble argument * Try some changes to bigint handling * Fix BigInt handling * Fix cls_longdouble test * Fix long double closure arg with no WASM_BIGINT * Use EM_JS to factor out js helpers * Support for varargs closure calls * Fix varargs calls * Fix err_bad_abi test * Fix typo in previous commit * Add more assertions to closures test suite * Fix some asserts * Add assertions to a few more tests * Fix some tests * Fix more floating point assertions * Update more tests * Var args for ffi_call * Don't do node tests * Macro for allocating on stack * Add some comments, simplify struct handling * Try again to fix varargs calls, add comments * Consolidate WASM_BIGINT conditionals into LOAD_U64 and STORE_U64 macros * A bit of cleanup * Fix another typo * Some fixes to the testsuite * Another testsuite fix * Fix varags with closures? * Another attempt at getting closure varargs to work * sig is initialized later * Allow libffi.closures tests to be run * Improve build script * Remove redundant semicolons * Fix a few libffi.closures test failures * Cleanup * Legacy dynCall API is no longer used * Fix FFI_TYPE_LONGDOUBLE offset * xfail 2 tests for WASM - closure_loc_fn0; not applicable -- codeloc doesn't point to closure. - huge_struct; function signature too long. * Revert some redundant dg-output/printf statements Helps Node. * Revert "Don't do node tests" This reverts commit a341ef4. * Fix assertions in cls_24byte * More tiny formating fixes to test suite * Revert "Revert "Don't do node tests"" This reverts commit 7722e68. * Fix 64 bit returns when WASM_BIGINT is absent * Fix print statement in cls_24byte * Add CALL_FUNC_PTR macro to allow pyodide to define custom calling behavior to handle fpcast * Update single_entry_structs tests * More explanations * Fix compile error in last commit * Add more support for pyodide fpcast emulation, update CI to try to test it * Clone via https * Fix path to pyodide emsdk_env * Add asserts to the rest of the test suite * Fix test compile errors * Fix some tests * Fix cls_ulonglong * Fix alignment of <4 byte args * fix cls_ulonglong again * Use snprintf instead of sprintf * Should assert than strncmp returned 0 * Fix va_struct1 and va_struct3 * Change double and long double tests These tests are failing because of a strange bug with prinft and doubles, but I am not convinced it necessarily has anything to do with libffi. This version casts the double to int before printing it and avoids the issue * Enable node tests * Revert "Change double and long double tests" This reverts commit 8f3ff89. * Fix PYODIDE_FPCAST flag * add conftest.py back in * Fix emcc error: setting `EXPORTED_FUNCTIONS` expects `<class 'list'>` but got `<class 'str'>` See discussion on pyodide/pyodide#1596 * Remove test.html * Remove duplicate test file * More changes from upstream * Fix some whitespace * Add some basic debug logging statements * Reapply libffi.exp changes * Don't build docs (#7) Works around build issue makeinfo: command not found. * Update long double alignment Emscripten 2.0.26 reduces the aligmnet of long double to 8. Quoting from `ChangeLog.md`: > The alignment of `long double`, which is a 128-bit floating-point > value implemented in software, is reduced from 16 to 8. The lower > alignment allows `max_align_t` to properly match the alignment we > use for malloc, which is 8 (raising malloc's alignment to achieve > correctness the other way would come with a performance regression). > (#10072) * Update long double alignment Emscripten 2.0.26 reduces the aligmnet of long double to 8. Quoting from `ChangeLog.md`: > The alignment of `long double`, which is a 128-bit floating-point > value implemented in software, is reduced from 16 to 8. The lower > alignment allows `max_align_t` to properly match the alignment we > use for malloc, which is 8 (raising malloc's alignment to achieve > correctness the other way would come with a performance regression). > (#10072) * Improve error handling a bit (#8) * Fix handling of signed arguments to ffi_call (#11) * Fix struct argument handling in ffi_call (#10) * Remove fpcast emulation tests * Align the stack to MAX_ALIGN before making call (#12) * Increase MAX_ARGS * Cleanup (#14) * Fix Closure compiler error with -sASSERTIONS=1 (#15) * Remove function pointer cast emulation (#13) This reverts commit 593b402 and cbc54da, as it's no longer needed after PR pyodide/pyodide#2019. * Prefer the `__EMSCRIPTEN__` definition over `EMSCRIPTEN` (#18) "The preprocessor define EMSCRIPTEN is deprecated. Don't pass it to code in strict mode. Code should use the define __EMSCRIPTEN__ instead." https://github.com/emscripten-core/emscripten/blob/84a634167a1cd9e8c47d37a559688153a4ceace6/emcc.py#L887-L890 * Install autoconf 2.71 * Try again with installing autoconf 2.71 * Fix compatibility with Emscripten 3.1.28 * CI: remove use of `EM_CONFIG` env See commit: emscripten-core/emsdk@3d87d5e * Fix cls_multi_schar: cast rest_call to signed char * Remove test xfails (#17) * Fix long double when used as a varargs argument * Enable unwindtest and fix it * Add EM_JS_DEPS * Also require convertJsFunctionToWasm * Run tests very very verbose * Echo the .emscripten file * Remove --experimental-wasm-bigint insertion * Build with assertions * Move verbosity flags back out of LDFLAGS * Remove debug print statement * Use up to date pyodide docker image * Explicitly cast res_call to fix test failure * Put back name of main function in cls_longdouble_va.c * Fix alignment The stack pointer apparently needs to be aligned to 16. There were some terrible subtle bugs caused by not respecting this. stackAlloc knows that the stack should be 16 aligned, so we can use stackAlloc(0) to enforce this. This way if alignment requirements change, as long as Emscripten updates stackAlloc to continue to enforce them we should be okay. * Fix handling of systems with no Js bigint integration When we run the node tests we use node v14 tests (since node v14 is vendored with Emscripten). Node v14 has no Js bigint integration unless the --experimental-wasm-bigint flag is passed. So only the node tests really notice if we get this right. Turns out, it didn't work. We can't call a JavaScript function with 64 bit integer arguments without bigint integration. In ffi_call, we are trying to call a wasm function that takes 64 bit integer arguments. dynCall is designed to do this. We need to go back to tracking the signature when we don't have WASM_BIGINT, and then use dynCall. This works better now that emscripten can dynamically fill in extra dynCall wrappers: emscripten-core/emscripten#17328 On the other hand, for the closures we are not getting a function pointer as a first argument. We need to make our own wasm legalizer adaptor that splits 64 bit integer arguments and then calls the JavaScript trampoline, then the JavaScript trampoline reassembles them, calls the closure, then splits the result (if it's a 64 bit integer) and the adaptor puts it back together. * Improvements to emscripten test shell scripts (#21) This fixes the C++ unwinding tests and makes other minor improvements to the Emscripten test shell scripts. * Rename the test folder and move test files into emscripten test folder * Use docker image that has autoconf-2.71 * Cleanup * Pin emscripten 3.1.30 * Fix build.sh path * Rearrange ci pipeline * Fix bpo_38748 test * Cleanup * Improvements to comments, add static asserts, and update copyright * Use `*_js` instead of `*_helper` for EM_JS functions (#22) * Minor code simplification * Xfail first dejagnu test to work around emscripten cache messages See emscripten-core/emscripten#18607 * Remove unneeded xfails * Shorten conftest.py by using pytest-pyodide * Apply formatters and linters to emscripten directory * Fix Emscripten xfail hack * Fix build-tests script * Patch emscripten to quiet info messages * Clean up compiler flags in scripts and remove some settings from circleci config * Rename emscripten quiet script * Add missing export * Don't remove go.exp * Add reference to emscripten logging issue --------- Co-authored-by: Kleis Auke Wolthuizen <[email protected]> Co-authored-by: Kleis Auke Wolthuizen <[email protected]> Co-authored-by: Christian Heimes <[email protected]>
* added build script * Apply libffi-emscripten patch * Some changes to wasm32/ffi.c * Remove exit(0); from test suites * Fix LONGDOUBLE argument type * Use more macros in ffi.c * Use switch statements instead of if chains * Implemented struct args * Finish struct implementation * Partially working closures * Got closures working (most of closures test suite passes) * Revert changes to test suite * Update .gitignore * Apply code formatter * Use stackSave and stackRestore rather than directly adjusting stack pointer * Add missing break * Fix visibility of ffi_closure_alloc and ffi_closure_free * Fix FFI_TYPE_STRUCT and FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used sig needs to be vi here for FFI_TYPE_STRUCT and FFI_TYPE_LONGDOUBLE, noticed this while running the test suite without WASM_BIGINT support. * Always use dynCall rather than direct wasmTable lookup (function pointer cast emulation changes dynCall) * Prevent closures.c from duplicating symbols * Try to set up CI * Add test with bigint * Make test methods static * Remove BigInt shorthand because it messes up terser * Add selenium tests * Update tests a bit to try to make CI work * WASM_BIGINT is a linker flag not a compile flag * Finish getting CI working (#1) * update gitignore * Avoid adding "use strict;" to generated JS This should be controlled by -s STRICT_JS in Emscripten. * Make JavaScript ES5 compliant * Remove redundant EXPORTED_RUNTIME_METHODS settings * Fix definition of DEREF_I16 * Avoid marshalling FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used * Add missing FFI_TYPE_STRUCT signature * Improve test scripts * Remove redundant EXPORTED_RUNTIME_METHODS settings * Add missing EOL * Add struct unpacking tests * Update ci config to try to actually use WASM_BIGINT * Revert "Avoid marshalling FFI_TYPE_LONGDOUBLE when WASM_BIGINT is not used" This reverts commit 61bd5a3e20891623715604581b6e872ab3dfab80. * Fix single_entry_structs tests * Fix return from closure call * Fix 64 bit return from closures * only allocate as much space on stack for return pointer as needed * Revert "only allocate as much space on stack for return pointer as needed" This reverts commit e54a30faea3803e7ac33eed191bde9e573850fc1. * xfail two tests * Fix err_bad_abi test * Remove test logging junk * Try to set up long double marshalling for closures * xfail err_bad_abi * Fix reference errors in previous commit * Add missing argument pointer assignment * Fix signature of function pointer in cls_dbls_struct * Fix longdouble argument * Try some changes to bigint handling * Fix BigInt handling * Fix cls_longdouble test * Fix long double closure arg with no WASM_BIGINT * Use EM_JS to factor out js helpers * Support for varargs closure calls * Fix varargs calls * Fix err_bad_abi test * Fix typo in previous commit * Add more assertions to closures test suite * Fix some asserts * Add assertions to a few more tests * Fix some tests * Fix more floating point assertions * Update more tests * Var args for ffi_call * Don't do node tests * Macro for allocating on stack * Add some comments, simplify struct handling * Try again to fix varargs calls, add comments * Consolidate WASM_BIGINT conditionals into LOAD_U64 and STORE_U64 macros * A bit of cleanup * Fix another typo * Some fixes to the testsuite * Another testsuite fix * Fix varags with closures? * Another attempt at getting closure varargs to work * sig is initialized later * Allow libffi.closures tests to be run * Improve build script * Remove redundant semicolons * Fix a few libffi.closures test failures * Cleanup * Legacy dynCall API is no longer used * Fix FFI_TYPE_LONGDOUBLE offset * xfail 2 tests for WASM - closure_loc_fn0; not applicable -- codeloc doesn't point to closure. - huge_struct; function signature too long. * Revert some redundant dg-output/printf statements Helps Node. * Revert "Don't do node tests" This reverts commit a341ef4b. * Fix assertions in cls_24byte * More tiny formating fixes to test suite * Revert "Revert "Don't do node tests"" This reverts commit 7722e685ea04e2420e042886816d8c4dd31f5dcb. * Fix 64 bit returns when WASM_BIGINT is absent * Fix print statement in cls_24byte * Add CALL_FUNC_PTR macro to allow pyodide to define custom calling behavior to handle fpcast * Update single_entry_structs tests * More explanations * Fix compile error in last commit * Add more support for pyodide fpcast emulation, update CI to try to test it * Clone via https * Fix path to pyodide emsdk_env * Add asserts to the rest of the test suite * Fix test compile errors * Fix some tests * Fix cls_ulonglong * Fix alignment of <4 byte args * fix cls_ulonglong again * Use snprintf instead of sprintf * Should assert than strncmp returned 0 * Fix va_struct1 and va_struct3 * Change double and long double tests These tests are failing because of a strange bug with prinft and doubles, but I am not convinced it necessarily has anything to do with libffi. This version casts the double to int before printing it and avoids the issue * Enable node tests * Revert "Change double and long double tests" This reverts commit 8f3ff89c6577dc99564181cd9974f2f1ba21f1e9. * Fix PYODIDE_FPCAST flag * add conftest.py back in * Fix emcc error: setting `EXPORTED_FUNCTIONS` expects `<class 'list'>` but got `<class 'str'>` See discussion on pyodide/pyodide#1596 * Remove test.html * Remove duplicate test file * More changes from upstream * Fix some whitespace * Add some basic debug logging statements * Reapply libffi.exp changes * Don't build docs (#7) Works around build issue makeinfo: command not found. * Update long double alignment Emscripten 2.0.26 reduces the aligmnet of long double to 8. Quoting from `ChangeLog.md`: > The alignment of `long double`, which is a 128-bit floating-point > value implemented in software, is reduced from 16 to 8. The lower > alignment allows `max_align_t` to properly match the alignment we > use for malloc, which is 8 (raising malloc's alignment to achieve > correctness the other way would come with a performance regression). > (#10072) * Update long double alignment Emscripten 2.0.26 reduces the aligmnet of long double to 8. Quoting from `ChangeLog.md`: > The alignment of `long double`, which is a 128-bit floating-point > value implemented in software, is reduced from 16 to 8. The lower > alignment allows `max_align_t` to properly match the alignment we > use for malloc, which is 8 (raising malloc's alignment to achieve > correctness the other way would come with a performance regression). > (#10072) * Improve error handling a bit (#8) * Fix handling of signed arguments to ffi_call (#11) * Fix struct argument handling in ffi_call (#10) * Remove fpcast emulation tests * Align the stack to MAX_ALIGN before making call (#12) * Increase MAX_ARGS * Cleanup (#14) * Fix Closure compiler error with -sASSERTIONS=1 (#15) * Remove function pointer cast emulation (#13) This reverts commit 593b402 and cbc54da, as it's no longer needed after PR pyodide/pyodide#2019. * Prefer the `__EMSCRIPTEN__` definition over `EMSCRIPTEN` (#18) "The preprocessor define EMSCRIPTEN is deprecated. Don't pass it to code in strict mode. Code should use the define __EMSCRIPTEN__ instead." https://github.com/emscripten-core/emscripten/blob/84a634167a1cd9e8c47d37a559688153a4ceace6/emcc.py#L887-L890 * Install autoconf 2.71 * Try again with installing autoconf 2.71 * Fix compatibility with Emscripten 3.1.28 * CI: remove use of `EM_CONFIG` env See commit: emscripten-core/emsdk@3d87d5e * Fix cls_multi_schar: cast rest_call to signed char * Remove test xfails (#17) * Fix long double when used as a varargs argument * Enable unwindtest and fix it * Add EM_JS_DEPS * Also require convertJsFunctionToWasm * Run tests very very verbose * Echo the .emscripten file * Remove --experimental-wasm-bigint insertion * Build with assertions * Move verbosity flags back out of LDFLAGS * Remove debug print statement * Use up to date pyodide docker image * Explicitly cast res_call to fix test failure * Put back name of main function in cls_longdouble_va.c * Fix alignment The stack pointer apparently needs to be aligned to 16. There were some terrible subtle bugs caused by not respecting this. stackAlloc knows that the stack should be 16 aligned, so we can use stackAlloc(0) to enforce this. This way if alignment requirements change, as long as Emscripten updates stackAlloc to continue to enforce them we should be okay. * Fix handling of systems with no Js bigint integration When we run the node tests we use node v14 tests (since node v14 is vendored with Emscripten). Node v14 has no Js bigint integration unless the --experimental-wasm-bigint flag is passed. So only the node tests really notice if we get this right. Turns out, it didn't work. We can't call a JavaScript function with 64 bit integer arguments without bigint integration. In ffi_call, we are trying to call a wasm function that takes 64 bit integer arguments. dynCall is designed to do this. We need to go back to tracking the signature when we don't have WASM_BIGINT, and then use dynCall. This works better now that emscripten can dynamically fill in extra dynCall wrappers: emscripten-core/emscripten#17328 On the other hand, for the closures we are not getting a function pointer as a first argument. We need to make our own wasm legalizer adaptor that splits 64 bit integer arguments and then calls the JavaScript trampoline, then the JavaScript trampoline reassembles them, calls the closure, then splits the result (if it's a 64 bit integer) and the adaptor puts it back together. * Improvements to emscripten test shell scripts (#21) This fixes the C++ unwinding tests and makes other minor improvements to the Emscripten test shell scripts. * Rename the test folder and move test files into emscripten test folder * Use docker image that has autoconf-2.71 * Cleanup * Pin emscripten 3.1.30 * Fix build.sh path * Rearrange ci pipeline * Fix bpo_38748 test * Cleanup * Improvements to comments, add static asserts, and update copyright * Use `*_js` instead of `*_helper` for EM_JS functions (#22) * Minor code simplification * Xfail first dejagnu test to work around emscripten cache messages See emscripten-core/emscripten#18607 * Remove unneeded xfails * Shorten conftest.py by using pytest-pyodide * Apply formatters and linters to emscripten directory * Fix Emscripten xfail hack * Fix build-tests script * Patch emscripten to quiet info messages * Clean up compiler flags in scripts and remove some settings from circleci config * Rename emscripten quiet script * Add missing export * Don't remove go.exp * Add reference to emscripten logging issue --------- Co-authored-by: Kleis Auke Wolthuizen <[email protected]> Co-authored-by: Kleis Auke Wolthuizen <[email protected]> Co-authored-by: Christian Heimes <[email protected]>
4 participants
Elevator Pitch
Enable file system persistence with IndexedDB (
IDBFS), Node.js (NODEFS), blobs (WORKERFS), proxy (PROXYFS).Motivation
Continuing the discussion from #328, this would allow a pyodide instance to work more naturally with persistent files.
Some external use cases:
Some internal use cases:
Changes
-l*fs.jstoMakefileto restore all backendsconsidered other backends, but as this has some impact on build time and size without a clear use case, I have not included them for now, but an eventual API should probably plan for supporting them explicitlyhoistFS.mount,FS.mkdirandFS.syncandIDBFSup to thepyodideAPI?FSto thepyodidenamespace, pending a better name...add a convenience API in e.g.loadPyodide({mount: "/home/user1"})?sys.prefixFindings
MEMFS(which must exist) withFS.sync(true, (err)=>{...})and likewise back out withFS.sync(false, (err) => {...})Dead-ends
The image above was gotten by hacking
main.c, emulating a naive/broken approach for polling. A better path is being investigated!