When possible, load 8 chars and mask out the tag and \0 terminator.
~19% faster on a Z840, ~5% faster on a PixelC.
Change-Id: I12d4b7d86c92c887b00f5d2480d3ff05a7042df1
Reviewed-on: https://skia-review.googlesource.com/135624
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Klein <mtklein@google.com>
Use std::swap instead. It does not appear that any external user
specializes SkTSwap, but some may still use it. This removes all use in
Skia so that SkTSwap can later be removed in a smaller CL. After that
the <utility> include can be removed from SkTypes.h.
Change-Id: If03d4ee07dbecda961aa9f0dc34d171ef5168753
Reviewed-on: https://skia-review.googlesource.com/135578
Reviewed-by: Hal Canary <halcanary@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
We currently blow through string chars without checking for end-of-input.
Maybe we could avoid this upfront, when we locate the stop char: try to
determine if it's part of an unterminated string, fail immediately if
so. Figuring out if the tail is an unterminated string seems
intractable though (requires arbitrarily deep tail parsing).
That brings us to plan B:
* treat scope-closing tokens (} & ]) as string terminators
(we know end-of-input points to one of these for sure)
* adjust matchString() to check for end-of-input
Bug: oss-fuzz:8899
Change-Id: Ic0a88a405548e8724b76faca525099a7e7037341
Reviewed-on: https://skia-review.googlesource.com/135145
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
The input is not guaranteed to contain well-formed scopes, so it's not
sufficient to check for end-of-stream only when popping the top-level
scope -- we have to check on every scope pop.
Bug: oss-fuzz:8898
Change-Id: I7399a8872187ec6714672cac2ff8fc7fbf3c2dfe
Reviewed-on: https://skia-review.googlesource.com/135059
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
1) skip redundant \0 terminator => Value is always zero-initialized
2) skip storing a len record => strlen is cheap for short strings
Change-Id: I3c10c9b9cf6155b95124e2c0194c59e9531a7ca4
Reviewed-on: https://skia-review.googlesource.com/135049
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Pass an explicit input size instead of requiring a C string.
Thanks to mtklein's clever trick, this has no measurable perf impact.
TBR=
Change-Id: Ic8cb1dc75f4d0814e5b2c80038d1b8d3a7b072ab
Reviewed-on: https://skia-review.googlesource.com/134946
Reviewed-by: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
ASAN opines that a nullptr memcpy dest is undefined behavior, even when
n == 0. ASAN may be right.
This doesn't occur internally, in the parser, but can be triggered with
the DOM builder API (as do some tests currently).
We could say "don't do that", but if someone wants to build an empty
string/array/object, it's kind of awkward to force them to provide a
valid source pointer instead of simply e.g. Array(nullptr, 0).
So let's guard for this case to make ASAN happy.
Change-Id: If12e39f5eb8b273f22bbb0b5fce3321bf6482173
Reviewed-on: https://skia-review.googlesource.com/134944
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Pass an explicit input size instead of requiring a C string.
Thanks to mtklein's clever trick, this has no measurable perf impact.
Change-Id: I64f210a9f653a78b05ab6b58fa34479504aa35ff
Reviewed-on: https://skia-review.googlesource.com/134940
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
* move most common accessor methods to the header, for inlining
* drop the lazy type checking semantics in favor of explicit guarded/unguarded
conversions
* revisit the public class hierarchy to better constrain type-bound APIs
* expose public type factories and add tests
* drop the empty-vector optimization -- allocating an external size_t in these
uncommon cases is better than paying for a conditional on every access.
TBR=
Change-Id: Ic609bb74f12cad1756865a2489ad56c03ecc5494
Reviewed-on: https://skia-review.googlesource.com/134845
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
This reverts commit 03b68421ca.
Reason for revert: Broke Debian9 builds
Original change's description:
> [skjson] Implementation/API tweaks
>
> * move most common accessor methods to the header, for inlining
> * drop the lazy type checking semantics in favor of explicit guarded/unguarded
> conversions
> * revisit the public class hierarchy to better constrain type-bound APIs
> * expose public type factories and add tests
> * drop the empty-vector optimization -- allocating an external size_t in these
> uncommon cases is better than paying for a conditional on every access.
>
> Change-Id: I24a7c75db3aa8b12c740c77ac7df4af4e3a1dff8
> Reviewed-on: https://skia-review.googlesource.com/134610
> Commit-Queue: Florin Malita <fmalita@chromium.org>
> Reviewed-by: Mike Klein <mtklein@google.com>
TBR=mtklein@google.com,fmalita@chromium.org
Change-Id: I2c681ef5c8d5fc15508e58b4b0f6ab9491b7d76f
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/134880
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
* move most common accessor methods to the header, for inlining
* drop the lazy type checking semantics in favor of explicit guarded/unguarded
conversions
* revisit the public class hierarchy to better constrain type-bound APIs
* expose public type factories and add tests
* drop the empty-vector optimization -- allocating an external size_t in these
uncommon cases is better than paying for a conditional on every access.
Change-Id: I24a7c75db3aa8b12c740c77ac7df4af4e3a1dff8
Reviewed-on: https://skia-review.googlesource.com/134610
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Klein <mtklein@google.com>
This reverts commit 2a2f675926.
Reason for revert: this appears to be what is holding up the Chrome roll.
Original change's description:
> SkTypes: extract SkTo
>
> Change-Id: I8de790d5013db2105ad885fa2683303d7c250b09
> Reviewed-on: https://skia-review.googlesource.com/133620
> Reviewed-by: Mike Klein <mtklein@google.com>
TBR=mtklein@google.com,halcanary@google.com
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Change-Id: Iafd738aedfb679a23c061a51afe4b98a8d4cdfae
Reviewed-on: https://skia-review.googlesource.com/134504
Reviewed-by: Hal Canary <halcanary@google.com>
Commit-Queue: Hal Canary <halcanary@google.com>
Module-related functionality is not required in dm/bench at this point.
Let's disable module deps flattening for now. As an immediate benefit
we can reduce the deps visibility for some internal skjson targets.
Bug: skia:8061
Change-Id: I0e09208964ca94b519121207ba1b6225059ef573
Reviewed-on: https://skia-review.googlesource.com/133822
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
For now this is only wired to a bench and a couple of tests.
Local numbers, for a ~500KB BM "compressed" json:
micros bench
2456.54 json_rapidjson nonrendering
1192.38 json_skjson nonrendering
Change-Id: I7b3514f84c7c525d1787722c43ad6095e3692563
Reviewed-on: https://skia-review.googlesource.com/127622
Reviewed-by: Mike Klein <mtklein@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
This reverts commit ba2f82986f.
Reason for revert: This is going to be hard to maintain long-term. I've fixed the python script to handle this case: https://skia-review.googlesource.com/c/skia/+/131580
Original change's description:
> Rename sg tests and samples targets to fix Visual Studio name conflict
>
> GN was emitting projects for the SG versions of tests and samples, which
> conflicted with the top-level versions of those. This made the normal GN
> VS IDE files confusing, and the merged ones totally broken.
>
> Change-Id: Ifd048258d4358ed2852d5263b16278c5b8ebd9c4
> Reviewed-on: https://skia-review.googlesource.com/131391
> Commit-Queue: Brian Osman <brianosman@google.com>
> Commit-Queue: Jim Van Verth <jvanverth@google.com>
> Commit-Queue: Florin Malita <fmalita@chromium.org>
> Auto-Submit: Brian Osman <brianosman@google.com>
> Reviewed-by: Jim Van Verth <jvanverth@google.com>
> Reviewed-by: Florin Malita <fmalita@chromium.org>
TBR=jvanverth@google.com,brianosman@google.com,fmalita@chromium.org,fmalita@google.com
Change-Id: I25553fdcfd063c920635fe1bd71553b578d0b0c8
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/131600
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
The constness on null_provider causes:
"default initialization of an object of const type
'const NullResourceProvider' without a user-provided default constructor"
See https://stackoverflow.com/a/47368753 for more.
TL;DR; clang 3.8 and older is not happy
Bug: skia:
Change-Id: Icfc38680163c3bd4952c0e35551706cad8dbfee6
Reviewed-on: https://skia-review.googlesource.com/131521
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Florin Malita <fmalita@chromium.org>
GN was emitting projects for the SG versions of tests and samples, which
conflicted with the top-level versions of those. This made the normal GN
VS IDE files confusing, and the merged ones totally broken.
Change-Id: Ifd048258d4358ed2852d5263b16278c5b8ebd9c4
Reviewed-on: https://skia-review.googlesource.com/131391
Commit-Queue: Brian Osman <brianosman@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Auto-Submit: Brian Osman <brianosman@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Florin Malita <fmalita@chromium.org>
Most of the existing clients don't care about nested resources.
Change-Id: Ie7991dd25ebbd679b5b49e5624772c7e19e7ec79
Reviewed-on: https://skia-review.googlesource.com/131141
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Update the script to search for headers in both 'sources' and 'public'.
Change-Id: I195c6e3720f3d3d99dea04628388821a58fa791b
Reviewed-on: https://skia-review.googlesource.com/130823
Reviewed-by: Mike Klein <mtklein@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
* relocate all SkSG-related files under modules/sksg/
* fix various tidbits to make non-sksg builds possible
* drop obsolete SampleSGInval.cpp
Change-Id: I54e6c5bb1a09f45030fa8d607b3eb3f7cba78957
Reviewed-on: https://skia-review.googlesource.com/130025
Commit-Queue: Florin Malita <fmalita@chromium.org>
Reviewed-by: Mike Klein <mtklein@google.com>
