<feed xmlns='http://www.w3.org/2005/Atom'>
<title>suspect-devices/circuitpython/py/qstr.c, branch 4.1.0-beta.0</title>
<subtitle>CircuitPython - a Python implementation for teaching coding with microcontrollers</subtitle>
<id>https://git.suspectdevices.com/suspect-devices/circuitpython/atom?h=4.1.0-beta.0</id>
<link rel='self' href='https://git.suspectdevices.com/suspect-devices/circuitpython/atom?h=4.1.0-beta.0'/>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/'/>
<updated>2018-08-17T00:40:57+00:00</updated>
<entry>
<title>Compress all translated strings with Huffman coding.</title>
<updated>2018-08-17T00:40:57+00:00</updated>
<author>
<name>Scott Shawcroft</name>
<email>scott@tannewt.org</email>
</author>
<published>2018-08-16T01:32:37+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=de5a9d72dcdaacdd5048195cd5bab007f4b2baef'/>
<id>urn:sha1:de5a9d72dcdaacdd5048195cd5bab007f4b2baef</id>
<content type='text'>
This saves code space in builds which use link-time optimization.
The optimization drops the untranslated strings and replaces them
with a compressed_string_t struct. It can then be decompressed to
a c string.

Builds without LTO work as well but include both untranslated
strings and compressed strings.

This work could be expanded to include QSTRs and loaded strings if
a compress method is added to C. Its tracked in #531.
</content>
</entry>
<entry>
<title>Support internationalisation.</title>
<updated>2018-08-07T21:58:57+00:00</updated>
<author>
<name>Scott Shawcroft</name>
<email>scott@tannewt.org</email>
</author>
<published>2018-07-31T23:53:54+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=933add6cd833bb6ba6682ad2b6eb478871415ff5'/>
<id>urn:sha1:933add6cd833bb6ba6682ad2b6eb478871415ff5</id>
<content type='text'>
</content>
</entry>
<entry>
<title>WIP: after merge; before testing</title>
<updated>2018-07-11T20:45:30+00:00</updated>
<author>
<name>Dan Halbert</name>
<email>halbert@halwitz.org</email>
</author>
<published>2018-07-11T20:45:30+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=7c219600a246d8956d0b23ea3f5d125a820e6b6a'/>
<id>urn:sha1:7c219600a246d8956d0b23ea3f5d125a820e6b6a</id>
<content type='text'>
</content>
</entry>
<entry>
<title>Limit qstr pool size to reduce memory waste.</title>
<updated>2018-07-03T12:45:51+00:00</updated>
<author>
<name>Scott Shawcroft</name>
<email>scott@tannewt.org</email>
</author>
<published>2018-06-28T17:43:39+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=cced51cbd23aa732f0243fe555b4427ff46267b4'/>
<id>urn:sha1:cced51cbd23aa732f0243fe555b4427ff46267b4</id>
<content type='text'>
</content>
</entry>
<entry>
<title>Introduce a long lived section of the heap.</title>
<updated>2018-01-24T18:33:46+00:00</updated>
<author>
<name>Scott Shawcroft</name>
<email>scott.shawcroft@gmail.com</email>
</author>
<published>2018-01-24T00:22:05+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=416abe33ed44e44d8f75342731080217e123278b'/>
<id>urn:sha1:416abe33ed44e44d8f75342731080217e123278b</id>
<content type='text'>
This adapts the allocation process to start from either end of the heap
when searching for free space. The default behavior is identical to the
existing behavior where it starts with the lowest block and looks higher.
Now it can also look from the highest block and lower depending on the
long_lived parameter to gc_alloc. As the heap fills, the two sections may
overlap. When they overlap, a collect may be triggered in order to keep
the long lived section compact. However, free space is always eligable
for each type of allocation.

By starting from either of the end of the heap we have ability to separate
short lived objects from long lived ones. This separation reduces heap
fragmentation because long lived objects are easy to densely pack.

Most objects are short lived initially but may be made long lived when
they are referenced by a type or module. This involves copying the
memory and then letting the collect phase free the old portion.

QSTR pools and chunks are always long lived because they are never freed.

The reallocation, collection and free processes are largely unchanged. They
simply also maintain an index to the highest free block as well as the lowest.
These indices are used to speed up the allocation search until the next collect.

In practice, this change may slightly slow down import statements with the
benefit that memory is much less fragmented afterwards. For example, a test
import into a 20k heap that leaves ~6k free previously had the largest
continuous free space of ~400 bytes. After this change, the largest continuous
free space is over 3400 bytes.
</content>
</entry>
<entry>
<title>py/qstr: Rewrite find_qstr to make manifest that it returns a valid ptr.</title>
<updated>2017-11-29T06:01:39+00:00</updated>
<author>
<name>Damien George</name>
<email>damien.p.george@gmail.com</email>
</author>
<published>2017-11-29T05:58:27+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=8e323b8fa84277531685985327fb682761abd53d'/>
<id>urn:sha1:8e323b8fa84277531685985327fb682761abd53d</id>
<content type='text'>
So long as the input qstr identifier is valid (below the maximum number of
qstrs) the function will always return a valid pointer.  This patch
eliminates the "return 0" dead-code.
</content>
</entry>
<entry>
<title>py/compile: Use alloca instead of qstr_build when compiling import name.</title>
<updated>2017-11-01T02:16:16+00:00</updated>
<author>
<name>Damien George</name>
<email>damien.p.george@gmail.com</email>
</author>
<published>2017-11-01T02:16:16+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=487dbdb26748b86cf247600af187437310311145'/>
<id>urn:sha1:487dbdb26748b86cf247600af187437310311145</id>
<content type='text'>
The technique of using alloca is how dotted import names are composed in
mp_import_from and mp_builtin___import__, so use the same technique in the
compiler.  This puts less pressure on the heap (only the stack is used if
the qstr already exists, and if it doesn't exist then the standard qstr
block memory is used for the new qstr rather than a separate chunk of the
heap) and reduces overall code size.
</content>
</entry>
<entry>
<title>py: Add verbose debug compile-time flag MICROPY_DEBUG_VERBOSE.</title>
<updated>2017-08-15T01:53:36+00:00</updated>
<author>
<name>Stefan Naumann</name>
<email>me@stefannaumann.de</email>
</author>
<published>2017-07-24T16:55:14+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=ace9fb54053c29574bdf81ffacc5ddcf9d4b45d9'/>
<id>urn:sha1:ace9fb54053c29574bdf81ffacc5ddcf9d4b45d9</id>
<content type='text'>
It enables all the DEBUG_printf outputs in the py/ source code.
</content>
</entry>
<entry>
<title>all: Use the name MicroPython consistently in comments</title>
<updated>2017-07-31T08:35:40+00:00</updated>
<author>
<name>Alexander Steffen</name>
<email>devel.20.webmeister@spamgourmet.com</email>
</author>
<published>2017-06-30T07:22:17+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=55f33240f3d7051d4213629e92437a36f1fac50e'/>
<id>urn:sha1:55f33240f3d7051d4213629e92437a36f1fac50e</id>
<content type='text'>
There were several different spellings of MicroPython present in comments,
when there should be only one.
</content>
</entry>
<entry>
<title>py: Fix wrong assumption that m_renew will not move if shrinking</title>
<updated>2016-11-02T12:15:41+00:00</updated>
<author>
<name>Colin Hogben</name>
<email>colin@infinnovation.co.uk</email>
</author>
<published>2016-10-31T14:05:56+00:00</published>
<link rel='alternate' type='text/html' href='https://git.suspectdevices.com/suspect-devices/circuitpython/commit/?id=f9b6b37cf65c4f65c4ad461d439fbf624c0f10c1'/>
<id>urn:sha1:f9b6b37cf65c4f65c4ad461d439fbf624c0f10c1</id>
<content type='text'>
In both parse.c and qstr.c, an internal chunking allocator tidies up
by calling m_renew to shrink an allocated chunk to the size used, and
assumes that the chunk will not move.  However, when MICROPY_ENABLE_GC
is false, m_renew calls the system realloc, which does not guarantee
this behaviour.  Environments where realloc may return a different
pointer include:

(1) mbed-os with MBED_HEAP_STATS_ENABLED (which adds a wrapper around
malloc &amp; friends; this is where I was hit by the bug);

(2) valgrind on linux (how I diagnosed it).

The fix is to call m_renew_maybe with allow_move=false.
</content>
</entry>
</feed>
