This way complex runtime features (e.g. ftrace, symbol tables) can
always yoink zipos support. This is important now that apelink.com
automates embedding symbol tables for multiple cpus.
This change ports APE Loader to Linux AARCH64, so that Raspberry Pi
users can run programs like redbean, without the executable needing
to modify itself. Progress has also slipped into this change on the
issue of making progress better conforming to user expectations and
industry standards regarding which symbols we're allowed to declare
This change takes an entirely new approach to the incremental linking of
pkzip executables. The assets created by zipobj.com are now treated like
debug data. After a .com.dbg is compiled, fixupobj.com should be run, so
it can apply fixups to the offsets and move the zip directory to the end
of the file. Since debug data doesn't get objcopy'd, a new tool has been
introduced called zipcopy.com which should be run after objcopy whenever
a .com file is created. This is all automated by the `cosmocc` toolchain
which is rapidly becoming the new recommended approach.
This change also introduces the new C23 checked arithmetic macros.
- Work towards improving non-optimized build support
- Introduce MODE=zero which is -O0 without ASAN/UBSAN
- Use system GCC when ~/.cosmo.mk has USE_SYSTEM_TOOLCHAIN=1
- Have package.com check .privileged code doesn't call non-privileged
This change implements a new approach to function call logging, that's
based on the GCC flag: -fpatchable-function-entry. Read the commentary
in build/config.mk to learn how it works.
There's a new program named ape/ape-m1.c which will be used to build an
embeddable binary that can load ape and elf executables. The support is
mostly working so far, but still chasing down ABI issues.
This makes breaking changes to add underscores to many non-standard
function names provided by the c library. MODE=tiny is now tinier and
we now use smaller locks that are better for tiny apps in this mode.
Some headers have been renamed to be in the same folder as the build
package, so it'll be easier to know which build dependency is needed.
Certain old misguided interfaces have been removed. Intel intrinsics
headers are now listed in libc/isystem (but not in the amalgamation)
to help further improve open source compatibility. Header complexity
has also been reduced. Lastly, more shell scripts are now available.
This change restores the .symtab symbol table files in our flagship
programs (e.g. redbean.com, python.com) needed to show backtraces. This
also rolls back earlier changes to zip.com w.r.t. temp directories since
the right way to do it turned out to be the -b DIR flag.
This change also improves the performance of zip.com. It turned out
mmap() wasn't being used, because zip.com was assuming a 4096-byte
granularity, but cosmo requires 65536. There was also a chance to speed
up stdio scanning using the unlocked functions.
Landlock Make will no longer sandbox prerequisites that end with a
trailing slash. This means you can use use directory prerequisites
for detecting deleted files when using using globbing, without the
effect of unveiling the entire directory. When you do want make to
unveil directories, you can omit the trailing slash.
We now guarantee TMPDIR will be defined on a per build rule basis. It'll
be an absolute path. It'll be secure and unique. It'll be rm -rf'd after
the last shell script line in your build rule is executed. If $TMPDIR is
already defined, then it'll be created as a subdirectory of your $TMPDIR
and then replace the variable with the new definition. The Landlock Make
repository will be updated with examples shortly after this change which
shall be known as Landlock Make 1.1.1.
See #530
It turned out that specifying all SRCS and INCS as dependencies on the
pattern rules for all headers, caused `make` memory usage to skyrocket
from 40mb ot 160mb. This change also reduces the build graph another 4%.
This change also removes the futimens() call on the Landlock Make output
file workaround, since it caused problems with commands like fixupobj
which modify-in-place. It turns out if a file is opened for writing and
then no writes actually occur, then the modified time doesn't change.
- 10.5% reduction of o//depend dependency graph
- 8.8% reduction in latency of make command
- Fix issue with temporary file cleanup
There's a new -w option in compile.com that turns off the recent
Landlock output path workaround for "good commands" which do not
unlink() the output file like GNU tooling does.
Our new GNU Make unveil sandboxing appears to have zero overhead
in the grand scheme of things. Full builds are pretty fast since
the only thing that's actually slowed us down is probably libcxx
make -j16 MODE=rel
RL: took 85,732,063µs wall time
RL: ballooned to 323,612kb in size
RL: needed 828,560,521µs cpu (11% kernel)
RL: caused 39,080,670 page faults (99% memcpy)
RL: 350,073 context switches (72% consensual)
RL: performed 0 reads and 11,494,960 write i/o operations
pledge() and unveil() no longer consider ENOSYS to be an error.
These functions have also been added to Python's cosmo module.
This change also removes some WIN32 APIs and System Five magnums
which we're not using and it's doubtful anyone else would be too
- Write tests for cthreads
- Fix bugs in pe2.com tool
- Fix ASAN issue with GetDosEnviron()
- Consolidate the cthread header files
- Some code size optimizations for MODE=
- Attempted to squash a tls linker warning
- Attempted to get futexes working on FreeBSD
- Fix some minor issues in ar.com
- Have execve() look for `ape` command
- Rewrite NT paths using /c/ rather /??/c:/
- Replace broken GCC symlinks with .sym files
- Rewrite $PATH environment variables on startup
- Make $(APE_NO_MODIFY_SELF) the default bootloader
- Add all build command dependencies to build/bootstrap
- Get the repository mostly building from source on non-Linux
- Add GetCpuCount() API to redbean
- Add unix.gmtime() API to redbean
- Add unix.readlink() API to redbean
- Add unix.localtime() API to redbean
- Perfect the new redbean UNIX module APIs
- Integrate with Linux clock_gettime() vDSO
- Run Lua garbage collector when malloc() fails
- Fix another regression quirk with linenoise repl
- Fix GetProgramExecutableName() for systemwide installs
- Fix a build flake with test/libc/mem/test.mk SRCS list
Continuous Integration (via runit and runitd) is now re-enabled on win7
and win10. The `make test` command, which runs the tests on all systems
is now the fastest and most stable it's been since the project started.
UBSAN is now enabled in MODE=dbg in addition to ASAN. Many instances of
undefined behavior have been removed. Mostly things like passing a NULL
argument to memcpy(), which works fine with Cosmopolitan Libc, but that
doesn't prevents the compiler from being unhappy. There was an issue w/
GNU make where static analysis claims a sprintf() call can overflow. We
also now have nicer looking crash reports on Windows since uname should
now be supported and msys64 addr2line works reliably.
We defined `noinline` as an abbreviation for the longer version
`__attribute__((__noinline__))` which caused name clashes since
third party codebases often write it as `__attribute__((noinline))`.
This commit makes numerous refinements to cosmopolitan memory handling.
The default stack size has been reduced from 2mb to 128kb. A new macro
is now provided so you can easily reconfigure the stack size to be any
value you want. Work around the breaking change by adding to your main:
STATIC_STACK_SIZE(0x00200000); // 2mb stack
If you're not sure how much stack you need, then you can use:
STATIC_YOINK("stack_usage_logging");
After which you can `sort -nr o/$MODE/stack.log`. Based on the unit test
suite, nothing in the Cosmopolitan repository (except for Python) needs
a stack size greater than 30kb. There are also new macros for detecting
the size and address of the stack at runtime, e.g. GetStackAddr(). We
also now support sigaltstack() so if you want to see nice looking crash
reports whenever a stack overflow happens, you can put this in main():
ShowCrashReports();
Under `make MODE=dbg` and `make MODE=asan` the unit testing framework
will now automatically print backtraces of memory allocations when
things like memory leaks happen. Bugs are now fixed in ASAN global
variable overrun detection. The memtrack and asan runtimes also handle
edge cases now. The new tools helped to identify a few memory leaks,
which are fixed by this change.
This change should fix an issue reported in #288 with ARG_MAX limits.
Fixing this doubled the performance of MKDEPS.COM and AR.COM yet again.
- Python static hello world now 1.8mb
- Python static fully loaded now 10mb
- Python HTTPS client now uses MbedTLS
- Python REPL now completes import stmts
- Increase stack size for Python for now
- Begin synthesizing posixpath and ntpath
- Restore Python \N{UNICODE NAME} support
- Restore Python NFKD symbol normalization
- Add optimized code path for Intel SHA-NI
- Get more Python unit tests passing faster
- Get Python help() pagination working on NT
- Python hashlib now supports MbedTLS PBKDF2
- Make memcpy/memmove/memcmp/bcmp/etc. faster
- Add Mersenne Twister and Vigna to LIBC_RAND
- Provide privileged __printf() for error code
- Fix zipos opendir() so that it reports ENOTDIR
- Add basic chmod() implementation for Windows NT
- Add Cosmo's best functions to Python cosmo module
- Pin function trace indent depth to that of caller
- Show memory diagram on invalid access in MODE=dbg
- Differentiate stack overflow on crash in MODE=dbg
- Add stb_truetype and tools for analyzing font files
- Upgrade to UNICODE 13 and reduce its binary footprint
- COMPILE.COM now logs resource usage of build commands
- Start implementing basic poll() support on bare metal
- Set getauxval(AT_EXECFN) to GetModuleFileName() on NT
- Add descriptions to strerror() in non-TINY build modes
- Add COUNTBRANCH() macro to help with micro-optimizations
- Make error / backtrace / asan / memory code more unbreakable
- Add fast perfect C implementation of μ-Law and a-Law audio codecs
- Make strtol() functions consistent with other libc implementations
- Improve Linenoise implementation (see also github.com/jart/bestline)
- COMPILE.COM now suppresses stdout/stderr of successful build commands
The ZIP filesystem has a breaking change. You now need to use /zip/ to
open() / opendir() / etc. assets within the ZIP structure of your APE
binary, instead of the previous convention of using zip: or zip! URIs.
This is needed because Python likes to use absolute paths, and having
ZIP paths encoded like URIs simply broke too many things.
Many more system calls have been updated to be able to operate on ZIP
files and file descriptors. In particular fcntl() and ioctl() since
Python would do things like ask if a ZIP file is a terminal and get
confused when the old implementation mistakenly said yes, because the
fastest way to guarantee native file descriptors is to dup(2). This
change also improves the async signal safety of zipos and ensures it
doesn't maintain any open file descriptors beyond that which the user
has opened.
This change makes a lot of progress towards adding magic numbers that
are specific to platforms other than Linux. The philosophy here is that,
if you use an operating system like FreeBSD, then you should be able to
take advantage of FreeBSD exclusive features, even if we don't polyfill
them on other platforms. For example, you can now open() a file with the
O_VERIFY flag. If your program runs on other platforms, then Cosmo will
automatically set O_VERIFY to zero. This lets you safely use it without
the need for #ifdef or ifstatements which detract from readability.
One of the blindspots of the ASAN memory hardening we use to offer Rust
like assurances has always been that memory passed to the kernel via
system calls (e.g. writev) can't be checked automatically since the
kernel wasn't built with MODE=asan. This change makes more progress
ensuring that each system call will verify the soundness of memory
before it's passed to the kernel. The code for doing these checks is
fast, particularly for buffers, where it can verify 64 bytes a cycle.
- Correct O_LOOP definition on NT
- Introduce program_executable_name
- Add ASAN guards to more system calls
- Improve termios compatibility with BSDs
- Fix bug in Windows auxiliary value encoding
- Add BSD and XNU specific errnos and open flags
- Add check to ensure build doesn't talk to internet
Thanks to all the refactorings we now have the ability to enforce
reasonable limitations on the amount of resources any individual
compile or test can consume. Those limits are currently:
- `-C 8` seconds of 3.1ghz CPU time
- `-M 256mebibytes` of virtual memory
- `-F 100megabyte` limit on file size
Only one file currently needs to exceed these limits:
o/$(MODE)/third_party/python/Objects/unicodeobject.o: \
QUOTA += -C16 # overrides cpu limit to 16 seconds
This change introduces a new sizetol() function to LIBC_FMT for parsing
byte or bit size strings with Si unit suffixes. Functions like atoi()
have been rewritten too.
- Abort if .init.lua fails
- Refactor redbean to use new append library
- Use first certificate if SNI routing fails
- Use function/data sections when building Lua
- Don't use self-signed auto-generated cert for client
- Add -D staging dirs to redbean lua module default path
redbean will now set Referer-Policy to no-referrer-when-downgrade on
text/html responses by default. There's better explanations on the bits
of security redbean is offering. In short, it's 128+ for modern clients
and 112+ for legacy. If the -B flag is used then it's 192+ for modern
and 150+ for non-EC.
One of the disadvantages of x25519 and ℘256 is it only provides 126 bits
of security, so that seems like a weak link in the chain, if we're using
ECDHE-ECDSA-AES256-GCM-SHA384. The U.S. government wants classified data
to be encrypted using a curve at least as strong as ℘384, which provides
192 bits of security, but if you read the consensus of stack exchange it
would give you the impression that ℘384 is three times slower.
This change (as well as the previous one) makes ℘384 three times as fast
by tuning its modulus and multiplication subroutines with new tests that
should convincingly show: the optimized code behaves the same way as the
old code. Some of the diff noise from the previous change is now removed
too, so that our vendored fork can be more easily compared with upstream
sources. So you can now have stronger cryptography without compromises.
℘384 modulus Justine l: 28𝑐 9𝑛𝑠
℘384 modulus MbedTLS NIST l: 127𝑐 41𝑛𝑠
℘384 modulus MbedTLS MPI l: 1,850𝑐 597𝑛𝑠
The benchmarks above show the improvements made by secp384r1() which is
an important function since it needs to be called 13,000 times whenever
someone establishes a connection to your web server. The same's true of
Mul6x6Adx() which is able to multiply 384-bit numbers in 73 cycles, but
only if your CPU was purchased after 2014 when Broadwell was introduced
This change makes SSL virtual hosting possible. You can now load
multiple certificates for multiple domains and redbean will just
figure out which one to use, even if you only have 1 ip address.
You can also use a jumbo certificate that lists all your domains
in the the subject alternative names.
This change also makes performance improvements to MbedTLS. Here
are some benchmarks vs. cc1920749e
BEFORE AFTER (microsecs)
suite_ssl.com 2512881 191738 13.11x faster
suite_pkparse.com 36291 3295 11.01x faster
suite_x509parse.com 854669 120293 7.10x faster
suite_pkwrite.com 6549 1265 5.18x faster
suite_ecdsa.com 53347 18778 2.84x faster
suite_pk.com 49051 18717 2.62x faster
suite_ecdh.com 19535 9502 2.06x faster
suite_shax.com 15848 7965 1.99x faster
suite_rsa.com 353257 184828 1.91x faster
suite_x509write.com 162646 85733 1.90x faster
suite_ecp.com 20503 11050 1.86x faster
suite_hmac_drbg.no_reseed.com 19528 11417 1.71x faster
suite_hmac_drbg.nopr.com 12460 8010 1.56x faster
suite_mpi.com 687124 442661 1.55x faster
suite_hmac_drbg.pr.com 11890 7752 1.53x faster
There aren't any special tricks to the performance imporvements.
It's mostly due to code cleanup, assembly and intel instructions
like mulx, adox, and adcx.
