profile
viewpoint
Giovanni Bajo rasky Develer S.r.l. Florence, Italy http://giovanni.bajo.it Developer Relation @italia / @teamdigitale. CTO of @develersrl and @greenapes; Python, Go, C++. Interested in algorithms & security.

rasky/gcs 84

Golomb Coded Sets

rasky/geventconnpool 75

Generic TCP connection pool for gevent-based applications

rasky/genemu 22

Basic Genesis / MegaDrive Emulator

rasky/g64drive 6

Portable Linux/Mac tool for Retroactive 64drive

rasky/CryptoFaxPA 5

CryptoFaxPA

italia/publiccode-parser-go 4

A Go parser for publiccode.yml files

rasky/dnsmasq-dnssec 3

Implementation of a DNSSEC resolver in dnsmasq (WIP)

rasky/freemyfeed 2

Use password-protected URLs for feeds in Google Reader

rasky/develhack 1

DevelHack 2016 entry

DevelBoard/barebox 0

Barebox fork for DevelBoard – http://www.barebox.org

push eventrasky/r64emu

Giovanni Bajo

commit sha c4a5d815ce0c399df299a451b48390ea7cec014d

test: run golden test until test is finished

view details

Giovanni Bajo

commit sha f736a3e9f2559f510b983188d3e59ca39b39700a

rsp: fix bug in partial loads (lbv/lsv/llv/lqv) and update tests

view details

Giovanni Bajo

commit sha a7426cff3517f2c82f7d87c30a7db7a6b64ba0b7

rsp: implement LUV/SUV (with tests)

view details

Giovanni Bajo

commit sha b9ab558ecc260ad609baa7ebf715fa86864b7f9a

mips64/rsp: fix unaligned address access in SP

view details

Giovanni Bajo

commit sha cde870384ac505587f2a8fcb3e5e2f8bd06eda58

rsp: fix unaligned, wrapping and negative-offset memory accesses Fixes #8

view details

Giovanni Bajo

commit sha 98d928ad357ffdf2517d8a646c4d9ad92eda4b5a

dbg: avoid crash with unnamed windows

view details

Giovanni Bajo

commit sha 2af81d1007262e1a72ea7013313c54d165883400

rsp: correct DMA transfers with size not multiple of 8

view details

push time in 7 days

issue closedrasky/r64emu

RSP vector load/store offset is signed

From doc/rsp.md:

offset is an unsigned offset added to the value of the base register (with some scaling, depending on the actual instruction).

Treating this as unsigned breaks a lot of microcode.

closed time in 7 days

KieronJ

issue closedrasky/go-lzo

compression of files

Hi, I am using go language. And i want to compress files which are retrieved by ls command. I want to use rasky/go-lzo package. please guide me how can i do this?

closed time in 19 days

poojakarale
GollumEvent

issue closedrasky/go-lzo

Proposal to dual license the project

Can this project be dual licensed? perhaps in Apache 2.0 or MIT such that Apache 2.0 projects can link against this library?

Current LICENSE.gpl is too restrictive and incompatible with Apache 2.0 released projects.

closed time in 19 days

harshavardhana

issue commentrasky/go-lzo

Proposal to dual license the project

No I'm sorry, it's not possible to dual-license it since I've faithfully converted the original C source code which is GPL so I don't feel I can change the license.

harshavardhana

comment created time in 19 days

issue closedrasky/go-lzo

Go1.13 golang.org/cl/161477 will break Decompress1X runtime error handling

Upcoming Go 1.13 change https://golang.org/cl/161477 updates the runtime error strings for index out of range errors, which breaks the assumption that Decompress1X func has on it at https://github.com/rasky/go-lzo/blob/master/decompress.go#L160.

$ go version
$ go version devel +431b5c69ca Tue Apr 16 21:54:01 2019 +0000 darwin/amd64
$ go test
--- FAIL: TestDecompInlen (0.00s)
panic: runtime error: index out of range [0] with length 0 [recovered]
        panic: runtime error: index out of range [0] with length 0 [recovered]
        panic: runtime error: index out of range [0] with length 0

goroutine 20 [running]:
testing.tRunner.func1(0xc0000e8100)
        /Users/herbie/work/go113/go/src/testing/testing.go:830 +0x3a3
panic(0x115b440, 0xc00010a020)
        /Users/herbie/work/go113/go/src/runtime/panic.go:619 +0x1b2
github.com/rasky/go-lzo.Decompress1X.func1(0xc000106e70)
        /Users/herbie/work/google3-go/lzo/decompress.go:165 +0xfa
panic(0x115b440, 0xc00010a020)
        /Users/herbie/work/go113/go/src/runtime/panic.go:619 +0x1b2
github.com/rasky/go-lzo.(*reader).ReadU16(...)
        /Users/herbie/work/google3-go/lzo/decompress.go:94
github.com/rasky/go-lzo.Decompress1X(0x1199c40, 0xc0000c2040, 0x31, 0x0, 0xc0000d6090, 0xf, 0x10, 0x0, 0x0)
        /Users/herbie/work/google3-go/lzo/decompress.go:234 +0x944
github.com/rasky/go-lzo.TestDecompInlen(0xc0000e8100)
        /Users/herbie/work/google3-go/lzo/compress_test.go:126 +0x20c
testing.tRunner(0xc0000e8100, 0x1177800)
        /Users/herbie/work/go113/go/src/testing/testing.go:865 +0xbf
created by testing.(*T).Run
        /Users/herbie/work/go113/go/src/testing/testing.go:916 +0x350
exit status 2
FAIL    github.com/rasky/go-lzo 0.024s

I'll send a PR to update the check.

closed time in 19 days

cybrcodr

issue commentrasky/go-lzo

Go1.13 golang.org/cl/161477 will break Decompress1X runtime error handling

I've merged the PR, sorry for the delay, I had missed it.

cybrcodr

comment created time in 19 days

pull request commentrasky/go-lzo

Update Decompress1X handling of runtime error for Go1.13.

Sorry, for some reason I had missed this PR.

cybrcodr

comment created time in 19 days

push eventrasky/go-lzo

Herbie Ong

commit sha 6da3e7744dae85473fb3df10c86e537ba07727c4

Update Decompress1X handling of runtime error for Go1.13. Upcoming Go 1.13 change https://golang.org/cl/161477 updates the runtime error strings for index out of range errors. This change simply makes the comparison a bit more loose to handle both pre-1.13 and 1.13.

view details

Giovanni Bajo

commit sha 96a758eda86e0ca29bc2638000fdb0ff0a6d96d8

Merge pull request #8 from cybrcodr/go113 Update Decompress1X handling of runtime error for Go1.13.

view details

push time in 19 days

PR merged rasky/go-lzo

Update Decompress1X handling of runtime error for Go1.13.

Upcoming Go 1.13 change https://golang.org/cl/161477 updates the runtime error strings for index out of range errors.

This change simply makes the comparison a bit more loose to handle both pre-1.13 and 1.13.

+2 -1

2 comments

1 changed file

cybrcodr

pr closed time in 19 days

issue commentgolang/go

cmd/compile: prove misses obvious facts about constants

Can you try if my CL https://go-review.googlesource.com/c/go/+/196679 fixes the test case you have in mind? It does add bounds to constants

josharian

comment created time in 24 days

push eventrasky/libdragon

Giovanni Bajo

commit sha f0b06439f42b55b27fde1568ef8dbf94543dbca7

Reimplement mtc2/mfc2 with the extended syntax allowed by RSP

view details

push time in a month

PR opened DragonMinded/libdragon

Improve error message when using MIPS opcodes not available on RSP

When using a MIPS opcode not available on RSP, before the user would get:

Error: unrecognized opcode 'swr_na'

After this PR, the error is more clear:

Error: swr is not supported by RSP

+1 -1

0 comment

1 changed file

pr created time in a month

push eventrasky/libdragon

Giovanni Bajo

commit sha e4e82b0cf0e4691bee4fc44a4107f6f8e369c50f

Improve error message when using MIPS opcodes not available on RSP

view details

push time in a month

create barnchrasky/libdragon

branch : notimplerror

created branch time in a month

push eventrasky/libdragon

Giovanni Bajo

commit sha 8eb82c1860a0b05a21047cd8ebe82a466bd00386

Reimplement mtc2/mfc2 with the extended syntax allowed by RSP

view details

push time in a month

PR opened DragonMinded/libdragon

Reimplement mtc2/mfc2 with the extended syntax allowed by RSP

The standard mfc2/mtc2 (as defined by MIPS architecture and GCC) do not allow to specify the "element" argument for the vector register. RSP allows that, by using unused space in the opcode.

+47 -0

0 comment

1 changed file

pr created time in a month

create barnchrasky/libdragon

branch : mtc2

created branch time in a month

fork rasky/mips-syntax

MIPS syntax highlightning package for sublime text 2

fork in a month

push eventdevelersrl/showtime2

Apruzzese Francesco

commit sha 701026f45ee62d090556e79a6c711caa1e0f0763

Switch to Odoo * [IMP] Python3 code * [FIX] Reference based on Odoo fields

view details

push time in a month

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

I also want to provide an additional datapoint: while it's true that using the docking feature for end-developers is very easy (just flip the docking flag), using the multi-viewport support requires more extensive changes in the backend plugins as the internal API for backend plugins is extended to support API to create multiple windows.

One of the goal of this PR is to help starting adapting these backends. For instance, I use rust-imgui-sdl2 that will have to be adapted. It's easier to do once there is a docking branch in imgui-rs, that can be tracked to verify progress and to check that the implementation is working.

Once DearImgui merges the main docking branch, ideally all the binding chain should just merge their own docking branch (cimgui -> imgui-rs -> rust-imgui-sdl2).

rasky

comment created time in a month

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

I personally disagree on point 2. For me, the best course of action is to use the standard cimgui upstream that already tracks the docking branch in their own docking branch, which is what my PR is based upon (one of the thing that my PR does is to change the submodule to track the docking branch).

Then, I'd personally prefer a docking branch in imgui-rs as well; the feature flag would be more difficult to maintain and I'm not sure it gives a big benefit, especially once you think that in some time there will not be two branches anymore but just one.

rasky

comment created time in a month

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

@Gekkio since you say that this PR has little information, what information do you need? The changes in this PR are simply “translations” from C++ to rust, adjustment of data structure to match the layout in C++, constants, and stuff like that: whatever is required to make tests pass. I didn’t write any new API (yet) so there’s zero design choices. I’m happy to provide missing information.

rasky

comment created time in a month

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

I hope there’s no need to fork, Rust community is small enough without having the need to further split maintenance of packages. @Gekkio I hope you will allow a co-maintainer of this very useful crate if you don’t have time to follow it.

rasky

comment created time in a month

Pull request review commentdevelersrl/showtime2

Openforce improvements

 def login(self, odoouri, user, password, db):             password=self.password)      def projects(self, client):-        project_model = client.AccountAnalyticAccount-        ids = project_model.search(['use_timesheets=True', 'invoice_on_timesheets=True', 'state=("open","pending")'])-        projects = project_model.read(ids, ['name'])+        project_model = client.ProjectProject+        project_ids = project_model.search([+            ('allow_timesheets', '=', True),+            ('billable_type', '!=', 'no'),+            ])+        projects = project_model.read(project_ids, ['name'])         return projects      def hours(self, client, projectids, from_date=None, to_date=None):-        # Read all billable types and differentiate between billable and non-billable-        # using the discount factor (if < 100.0, we consider it billable, since 100%-        # is non billable).-        billable = set()-        for fact in client.Hr_timesheet_invoiceFactor.read([]):-            if fact["factor"] < 100:-                billable.add(fact["id"])--        timesheet_model = client.HrAnalyticTimesheet+        timesheet_model = client.AccountAnalyticLine         ids = timesheet_model.search([-            'account_name=%s' % ",".join(['"'+p+'"' for p in projectids]),-            'date >= %s' % from_date,-            'date < %s' % to_date,+            ('project_id', 'in', projectids),+            ('date', '>=', from_date.strftime('%Y-%m-%d')),+            ('date', '<', to_date.strftime('%Y-%m-%d')),+            ('timesheet_invoice_type', '!=', 'non_billable_project'),         ])         if not ids:             return [] -        data = timesheet_model.read(ids, ['account_name', 'user_id', 'date', 'line_id', 'to_invoice', 'unit_amount'])-+        data = timesheet_model.read(ids, ['project_id', 'user_id', 'date',+                                          'move_id', 'timesheet_invoice_type',+                                          'unit_amount'])         hours = []         for item in data:-            if not item['to_invoice']:-                continue             h = {}-            h["project"] = item["account_name"]+            h["project"] = item["project_id"]             h["date"] = item["date"]             h["user"] = item["user_id"][1]-            h["remark"] = item["line_id"][1]+            h["remark"] = item["move_id"][1] if item['move_id'] else ''

@OpenCode ti torna tutto? Posso procedere al merge o devi cambiare qualcosa vista la mia analisi?

OpenCode

comment created time in a month

Pull request review commentdevelersrl/showtime2

Openforce improvements

 def login(self, odoouri, user, password, db):             password=self.password)      def projects(self, client):-        project_model = client.AccountAnalyticAccount-        ids = project_model.search(['use_timesheets=True', 'invoice_on_timesheets=True', 'state=("open","pending")'])-        projects = project_model.read(ids, ['name'])+        project_model = client.ProjectProject+        project_ids = project_model.search([+            ('allow_timesheets', '=', True),+            ('billable_type', '!=', 'no'),+            ])+        projects = project_model.read(project_ids, ['name'])         return projects      def hours(self, client, projectids, from_date=None, to_date=None):-        # Read all billable types and differentiate between billable and non-billable-        # using the discount factor (if < 100.0, we consider it billable, since 100%-        # is non billable).-        billable = set()-        for fact in client.Hr_timesheet_invoiceFactor.read([]):-            if fact["factor"] < 100:-                billable.add(fact["id"])--        timesheet_model = client.HrAnalyticTimesheet+        timesheet_model = client.AccountAnalyticLine         ids = timesheet_model.search([-            'account_name=%s' % ",".join(['"'+p+'"' for p in projectids]),-            'date >= %s' % from_date,-            'date < %s' % to_date,+            ('project_id', 'in', projectids),+            ('date', '>=', from_date.strftime('%Y-%m-%d')),+            ('date', '<', to_date.strftime('%Y-%m-%d')),+            ('timesheet_invoice_type', '!=', 'non_billable_project'),         ])         if not ids:             return [] -        data = timesheet_model.read(ids, ['account_name', 'user_id', 'date', 'line_id', 'to_invoice', 'unit_amount'])-+        data = timesheet_model.read(ids, ['project_id', 'user_id', 'date',+                                          'move_id', 'timesheet_invoice_type',+                                          'unit_amount'])         hours = []         for item in data:-            if not item['to_invoice']:-                continue             h = {}-            h["project"] = item["account_name"]+            h["project"] = item["project_id"]             h["date"] = item["date"]             h["user"] = item["user_id"][1]-            h["remark"] = item["line_id"][1]+            h["remark"] = item["move_id"][1] if item['move_id'] else ''

Ah noto ora che la stessa descrizione è presente banalmente in name. Non so se all'epoca non l'avevo vista o non c'era. Per me è indifferente ovviamente, basta estrarre quel campo dovunque esso sia.

OpenCode

comment created time in a month

Pull request review commentdevelersrl/showtime2

Openforce improvements

 def login(self, odoouri, user, password, db):             password=self.password)      def projects(self, client):-        project_model = client.AccountAnalyticAccount-        ids = project_model.search(['use_timesheets=True', 'invoice_on_timesheets=True', 'state=("open","pending")'])-        projects = project_model.read(ids, ['name'])+        project_model = client.ProjectProject+        project_ids = project_model.search([+            ('allow_timesheets', '=', True),+            ('billable_type', '!=', 'no'),+            ])+        projects = project_model.read(project_ids, ['name'])         return projects      def hours(self, client, projectids, from_date=None, to_date=None):-        # Read all billable types and differentiate between billable and non-billable-        # using the discount factor (if < 100.0, we consider it billable, since 100%-        # is non billable).-        billable = set()-        for fact in client.Hr_timesheet_invoiceFactor.read([]):-            if fact["factor"] < 100:-                billable.add(fact["id"])--        timesheet_model = client.HrAnalyticTimesheet+        timesheet_model = client.AccountAnalyticLine         ids = timesheet_model.search([-            'account_name=%s' % ",".join(['"'+p+'"' for p in projectids]),-            'date >= %s' % from_date,-            'date < %s' % to_date,+            ('project_id', 'in', projectids),+            ('date', '>=', from_date.strftime('%Y-%m-%d')),+            ('date', '<', to_date.strftime('%Y-%m-%d')),+            ('timesheet_invoice_type', '!=', 'non_billable_project'),         ])         if not ids:             return [] -        data = timesheet_model.read(ids, ['account_name', 'user_id', 'date', 'line_id', 'to_invoice', 'unit_amount'])-+        data = timesheet_model.read(ids, ['project_id', 'user_id', 'date',+                                          'move_id', 'timesheet_invoice_type',+                                          'unit_amount'])         hours = []         for item in data:-            if not item['to_invoice']:-                continue             h = {}-            h["project"] = item["account_name"]+            h["project"] = item["project_id"]             h["date"] = item["date"]             h["user"] = item["user_id"][1]-            h["remark"] = item["line_id"][1]+            h["remark"] = item["move_id"][1] if item['move_id'] else ''

Questo è un record di quella tabella:

{'account_id': [833,
                'Progetti /  Unitec 2016 07 HW - Industrializzazione 4 M6'],
 'account_name': ' Unitec 2016 07 HW - Industrializzazione 4 M6',
 'amount': -15.5,
 'amount_currency': 0.0,
 'code': False,
 'company_id': [1, 'Develer SRL'],
 'currency_id': False,
 'date': '2019-12-04',
 'date_aal': '2019-12-04',
 'general_account_id': [1069, '72 05 010 Salari E Stipendi'],
 'id': 430501,
 'invoice_id': False,
 'journal_id': [6, 'Timesheet'],
 'line_id': [783553, 'Organizzazione collaudo'],
 'move_id': False,
 'name': 'Organizzazione collaudo',
 'partner_id': [5687, 'Unitec Spa'],
 'product_id': [3425, 'Sviluppo software - Nicola Sansonne'],
 'product_uom_id': [12, 'Ore'],
 'ref': False,
 'sheet_id': [3000, 'Month 12'],
 'to_invoice': [11, 'Non Billable'],
 'unit_amount': 0.5,
 'user_id': [104, 'Sansonne  Nicola']}```

Io non ho molte altre informazioni se non che quel campo contiene la descrizione della registrazione ore effettuata. Cioè quando l'utente registra le ore lavorate, per ogni registrazione c'è una descrizione che può inserire, e in quel campo io la trovo. Più di questo non saprei...
OpenCode

comment created time in a month

issue commentgolang/go

proposal: allow && and || operators and parentheses in build tags

@ianlancetaylor when you say “we can’t remove”, I don’t understand whether you mean “we can’t disable support for the old syntax starting from Go 1.X for any X” or “we can’t remove the code parsing the old syntax from the toolchain ever”. I’m assuming you mean the former.

Why don’t we just try? It’s a relatively isolated corner of the language syntax, and the worst that can happen is that we make a giant mistake of the migration plan and we end up having two syntax for build tags forever. The best it can happen is that we learn if the Go ecosystem/community moves fast enough to allow a backward incompatible change to happen in a couple (?) of years. It’s a good story to write either case, and the worst case seems far less worse than doing the same mistake on other parts of the language.

I think this issue is a good “reality check” for the Go 2 migration plan. My feeling is that if this can’t be done, then Go 2 can’t succeed in changing mistakes of the past (maybe it will succeed in other areas).

mundaym

comment created time in a month

issue commentgolang/go

proposal: allow && and || operators and parentheses in build tags

I’m not sure I understand the rationale for declining. Everybody agrees that the current syntax is confusing. Go2 seems a good avenue to fix it for good. We also have a way to deprecate the current syntax (with go version checks) and move on to the new syntax for good (rather than having two syntaxes working at the same time). Also, it would be absolutely trivial to provide a go fix translation (heck, even go fmt could change it if we wanted do).

Why should we avoid changing just because the current one is “general enough”? The proposal itself was not even commented upon, and it provides a far superior syntax, much easier to read and understand.

If this proposal must be declined, i think it needs a stronger rationale than “the current syntax is general enough”.

mundaym

comment created time in a month

delete branch develersrl/sendtime

delete branch : openforceit-openforce_improvements

delete time in a month

push eventdevelersrl/sendtime

Giovanni Bajo

commit sha 9c7c3d1406b3c975793187d09daa1fe822745bc2

Odoo 12 and Python3 port * [ADD] Check route to have a response on working code * [FIX] timesheet create on Odoo 12.0 models * [IMP] Token generation based on python3 * [IMP] Create timesheet for specific user * [REF] Refactoring user information Odoo crypt password now so every time we try to get it, we have an empty string. With old code every user password would have been overwritten. With this refactoring we use an admin/manager account to write data for other users but force on creation the right user reference. * [ADD] README and sample setting file Co-authored-by: Apruzzese Francesco <cescoap@gmail.com>

view details

push time in a month

PullRequestEvent

PR closed develersrl/sendtime

Openforce improvements
+151 -105

2 comments

3 changed files

OpenCode

pr closed time in a month

pull request commentdevelersrl/sendtime

Openforce improvements

È sufficiente "spostarlo" in un branch, ci vuole due minuti. Lo faccio io ma chiudo questa PR perché viene dal tuo fork e lì non posso pushare.

OpenCode

comment created time in a month

create barnchdevelersrl/sendtime

branch : openforceit-openforce_improvements

created branch time in a month

push eventdevelersrl/sendtime

Francesco Apruzzese

commit sha 6e1303978e0c2ad2a8745b12a70ca71fc0eb67cf

[FIX] PEP8 code style

view details

push time in a month

MemberEvent
MemberEvent

Pull request review commentdevelersrl/sendtime

Openforce improvements

+DEBUG = True+SECRET_KEY = b'd54sade54wqa7c4r863ew7aqrx819734r8q47x634x2$43543'

Metti qualcosa tipo "CHANGEME" qui, vorrei evitare che ci siano token che sembrano veri nei file di esempio.

OpenCode

comment created time in 2 months

issue commentgolang/go

cmd/compile: teach prove about incrementing, len, and overflow

It should be fixed by the patch serie finishing at CL 196680. I can't test it right now though.

josharian

comment created time in 2 months

PR opened DragonMinded/libdragon

Fix vand opcode

The current value was a copy/paste error from previous line.

+1 -1

0 comment

1 changed file

pr created time in 2 months

push eventrasky/libdragon

Giovanni Bajo

commit sha 704df39e01262648e36b3bde22a306dc0344343d

Fix vand opcode

view details

push time in 2 months

issue openedanacierdem/libdragon-docker

Slim down image size

Currently, the image size is 1.33Gb. But it contains many things that can probably be pulled:

  • /usr/local/libexec/gcc/mips64-elf is not stripped (574Mb -> 56Mb)
  • /usr/local/bin is not stripped (110Mb -> 23Mb)
  • /usr/local/share/locale is 13Mb, not so big but do we need translation messages for GCC?
  • Running apt-get remove gcc && apt autoremove removes 200Mb of the native toolchain which is not needed.

created time in 2 months

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

Ping @Gekkio, any feedback is appreciated

rasky

comment created time in 2 months

push eventrasky/r64emu

Giovanni Bajo

commit sha 71b6ef758a27b5756edfec6d39e89adf2aeddf80

log: add convention for special keys with semantic meaning

view details

Giovanni Bajo

commit sha 5690b73884fa29e253a37a3b068c6c9304d5cc39

sync: adapt to the new convention of special keys for PC/FRAME

view details

Giovanni Bajo

commit sha 06da6012335092b6f92ee582012f6e7cdbd88680

dbg: improve logview by parsing special log keys, and optimize performance

view details

Giovanni Bajo

commit sha 2fd107acbd4525cc7d240388d71cb2a53902aed4

Split LogPool and move it into emu::log.

view details

Giovanni Bajo

commit sha 9693cf62bf4d562ec7d21ba35c57078a9e8cd77a

log: allow multiple log filters/views over a logpool

view details

Giovanni Bajo

commit sha 2fd5b0e5189ad9f2132e74adcc91229b4fc97191

dbg: fully implement multiple logview windows

view details

Giovanni Bajo

commit sha 6d3691598ac1c86f035a58ec07f5fc5277a7ff7e

dbg: implement log saving to disk

view details

Giovanni Bajo

commit sha 68330d558a457042c06ee35f363b0da2497358ef

dbg: in logview, add frame filter, line popup, and fix some off-by-one bugs in scrolling

view details

Giovanni Bajo

commit sha db15b80cdc31e60cd2f101ebb926db20ab599d7c

dbg: small refactoring on disasmview to expose force_pc

view details

Giovanni Bajo

commit sha de6fda73d1772c8e5db4982d277ee11eed7e9eda

dbg: display location data attached to loglines

view details

Giovanni Bajo

commit sha e9749532d85994752b4d48c7c1dd125f7d1eee37

dbg: improve logview kv popup with word wrapping of values

view details

Giovanni Bajo

commit sha 15792fc6f2189c1f807595f317b7deff52b4ab29

dbg: restore logview horizontal scrolling

view details

Giovanni Bajo

commit sha 576e4e33dff3d342b8f727724a40aab394a4abd0

dbg: fix logview shortcut "F" to only trigger when on focus

view details

Giovanni Bajo

commit sha 42ba4344ef40b364a98a0bed33e423bc2a609149

dbg: implement new memory view

view details

Giovanni Bajo

commit sha fc9e39a7b52760981487e127241611fbf74f1cae

mips64: expose memory view to debugger

view details

Giovanni Bajo

commit sha a92a97f907f01c791f1c37c8c3fc362a523f37c1

dbg: several small improvements to memory view

view details

push time in 2 months

pull request commentGekkio/imgui-rs

Add initial bindings for docking branch

Ping @Gekkio

rasky

comment created time in 2 months

issue commentGekkio/imgui-rs

Modal Popups

Currently, there's no way to set the size of a popup. It looks like I need to call the low-level igSetNextWindowSize. Probably, ModalPopup should grow a size method like other windows.

sanboxrunner

comment created time in 2 months

push eventrasky/r64emu

Giovanni Bajo

commit sha 204d1ad415e64f0bbd82803a4f8918da81cafc5f

Add new log view to debugger

view details

Giovanni Bajo

commit sha 993cca5d1ada9c60f894a3f96e368c3620b9c31a

dbg: further improvements to logview

view details

Giovanni Bajo

commit sha 7d4536f5362dc4ad1dd8d2590df132c6c47330d6

dbg: add tooltip with base conversion in all register inputs

view details

push time in 2 months

push eventrasky/r64emu

Giovanni Bajo

commit sha 534a902131ef7a1105e06facc7acb2a72d023e7e

Add support for highlighting in/out regs on current insn

view details

Giovanni Bajo

commit sha 9735ba039263b90b2efd8d17f4accdfac00f7d41

test: add missing call to LPV/SPV test

view details

Giovanni Bajo

commit sha cf1df7a6bcafb9df39812a500cc658c697fe3a95

rsp: add decoding for SPV/LPV

view details

Giovanni Bajo

commit sha 9ace6a263aad0c29dbef9ee4ca6550eb7ef4c253

Upgrade to imgui-rs 0.2 with breaking API changes

view details

Giovanni Bajo

commit sha 2945f55a2ad6c2247b15330cf67b986c6f833ca7

dbg: fix focusing check on disassembly views

view details

Giovanni Bajo

commit sha 3b8d5e46856e2d95776657e3cf8a0ef66244b5a4

mips64: mask access sizes during trace

view details

Giovanni Bajo

commit sha 5843681e31349f9edf7ff60f4cbd77250b3d8072

misp64: fix FPU registers display in 32-bit mode

view details

Giovanni Bajo

commit sha ea624fd9b342124a7af26963ddd6b3b8ec784696

sp: display VU registers in more compact form

view details

Giovanni Bajo

commit sha 8486e11f8d957a0149addf0f4124ee90710e9fb2

sp: unify VU register names

view details

Giovanni Bajo

commit sha b88f8bc021920ab7b1eb71bd7af5e41e503b7705

dbg: do not process global hotkeys while user is writing text into an input

view details

Giovanni Bajo

commit sha 5714fc2d0aefeb7c644c171f352bae986f52fe7e

dbg: allow click on blank spaces in disassembly view

view details

Giovanni Bajo

commit sha 5be9e8a8df2b9ed837cbd766f1475540fd120abb

dbg: allow decoding up to 8 args

view details

Giovanni Bajo

commit sha a2998470564ca2ae92819393ab5a72cc6d0c6a3f

sp: add accumulator as hidden output to VU multipications

view details

Giovanni Bajo

commit sha 3cf3ca81d941b4b4935b40f0382e92ff885eedf5

dbg: do not allow inserting too long text when editing registers

view details

Giovanni Bajo

commit sha a030f6738605e6e0ac19cc4c7013ad5fbeb43b55

dbg: fix Vec8x16 editing

view details

push time in 3 months

PR opened Gekkio/imgui-rs

Add initial bindings for docking branch

This PR adds initial support for the Dear Imgui docking branch.

The docking/viewport branch is a long-living branch that will eventually be merged to master and brings many new features to Dear Imgui. While some features can be used almost without changing code (docking), others require extensive support in the backend (viewports). Since it might take a while to correctly implement for viewports in our several backends, I thought we might as well start implementation.

This PR has initial support: it passes cargo test by fixing the layouts, but does not yet wrap new APIs. It was written off v0.2, as 0.3 is still not stable.

NOTE: This PR should probably be merged into a new docking branch. There's no way for me to target a branch that doesn't exist yet; if somebody creates a docking branch, I can change this PR and let it point there.

+2876 -3792

0 comment

9 changed files

pr created time in 3 months

push eventrasky/imgui-rs

Giovanni Bajo

commit sha ebbf09ef0282109c3833b8c4dce98cce4ec275e2

Add bindings for docking branch

view details

push time in 3 months

push eventrasky/imgui-rs

Giovanni Bajo

commit sha d502d4cecf6f36dd4e691a586ceb6eef2dd92174

Add bindings for docking branch

view details

push time in 3 months

create barnchrasky/imgui-rs

branch : docking

created branch time in 3 months

delete branch rasky/cimgui

delete branch : docking

delete time in 3 months

pull request commentcimgui/cimgui

Create a branch tracking the upstreaming docking branch

Thanks!

rasky

comment created time in 3 months

pull request commentcimgui/cimgui

Create a branch tracking the upstreaming docking branch

It should not be merged on máster branch but on docking branch instead.

Yes, my PR is meant to be merged on a new branch. I can’t do that from the Github UI, until you create a branch on this repo; then I can redo the PR on it.

Why not viewport branch instead of docking?

Viewport branch doesn’t exist anymore, now it’s part of the docking branch.

Was the generation done without any issues?

Yes.

If so, why is commit needed?

Because it lowers the barrier to adapt other bindings. For instance, after you merge this into a new branch, I can open a new PR on the rust bindings, which import cimgui as sub module; I can make the submodule track the docking branch here, and adapt the Rust bindings to export the new APIs (non-trivial); and after that is done, I have PRs for the rust-imgui-OpenGL backend, which in turn will have to point to a “docking” branch in Rust’s bindings, plus non-trivial additions in the code to support viewports. If we then update the branch here every month, the new features and bug fixes are easy to import.

If you don’t commit this, the rust bindings has no simple way of using the docking branch. I would have to talk the Rust maintainer to implement a docking branch that imports a forked version of cimgui that I would have to maintain elsewhere, and the Rust maintainer could easily object that basing work on a non-official fork of cimgui can mean wasting time. I would think the same, if I were the maintainer.

I hate long-lived branches with a passion, but that’s where we are. Docking branch is big and has many changes, the sooner we start supporting in non-C++ languages, the sooner we can benefit from it once it’s merged.

I understand that for cimgui there’s almost nothing to do, but in other languages’ back ends there’s lot of code to write, and this PR has no cost for you and helps downstream a lot. I can send regular updates via PRs to track upstream, and I can take care of maintaining it and raising issues I meet with you, if ever: those are all things that might end up on your plate anyway after it’s merged.

rasky

comment created time in 3 months

issue commentrasky/r64emu

RSP vector load/store offset is signed

Thanks! That’s a part of testing I’m still missing. I’ll hop on it

KieronJ

comment created time in 3 months

push eventrasky/cimgui

sonoro1234

commit sha d9e1d9a80d621cd96d9900ac340092853100416f

drop cimgui_internal.h generation

view details

Giovanni Bajo

commit sha f09a6a4208e6ccebf30f12188eb4ad8033c81c50

Regenerate bindings for docking branch

view details

push time in 3 months

PR opened cimgui/cimgui

Create a branch tracking the upstreaming docking branch

In preparation for the upcoming upstream docking/viewport branch merge, I think it would be very useful if the ecosystem of bindings could begin implementing support for it, since its a large change that has multiple effects also on backends (especially the multi viewport support).

This means that, in turn, it would be great if cimgui could publish itself a docking branch, tracking the upstream one, as that would make so much easier for other language bindings to follow. This PR implements support for it, to the best of my understanding. I didn't test it much at this point, but I managed to recompile a non-trivial application with it and it seems to work. Anyway, I think it's a first step.

+2706 -247

0 comment

14 changed files

pr created time in 3 months

create barnchrasky/cimgui

branch : docking

created branch time in 3 months

fork rasky/cimgui

c-api for imgui (https://github.com/ocornut/imgui)

fork in 3 months

pull request commentmichaelfairley/rust-imgui-opengl-renderer

Fix support for high-dpi displays

@michaelfairley ping on this, I've got the same bug

alexkirsz

comment created time in 3 months

push eventrasky/r64emu

Giovanni Bajo

commit sha 74c82c6b0e90c2efd7fbe613944f0fd9807ec90b

rsp: implement LPV/SPV

view details

push time in 3 months

push eventrasky/r64emu

Giovanni Bajo

commit sha 7c06579fd8dc05e6cdc59a8021d82c21ba9d5eeb

Add u64 support to testsuite

view details

Giovanni Bajo

commit sha 8950dcd4c0c711187aa9093a2e986a3b8aad77b3

Update to latest nighlty and do a general cargo update

view details

Giovanni Bajo

commit sha dc25e24aa79e239e3f7403ad657a8a9a063ac9ff

memint: make AccessSize serializable

view details

Giovanni Bajo

commit sha cebc4fde26f3fd2dd8f0e6f32a4a96b5f4188ab8

mips64: implement teq opcode

view details

Giovanni Bajo

commit sha ef3d000cf9fe7774ca27810a996b9d02a09ef214

mips64: fix typo in debug rendering for cop2/3

view details

Giovanni Bajo

commit sha 423fd661e837959da088c8cdb3219c7088d0cadc

mips64: more ops decoding

view details

Giovanni Bajo

commit sha cc50021ce671a621565060458d8ecb10e11a6973

mips64: implement DTMC1

view details

Giovanni Bajo

commit sha c9f24be0765b7c422a11cd123f2dd3032963eeb2

emu: add wcb option for memory banks

view details

Giovanni Bajo

commit sha 12fe88b77e16b68174bc7bac72041202eb15bdac

Add automatic load/save of breakpoints/watchpoints

view details

Giovanni Bajo

commit sha 0a10d9a89e7ee30b8d15644eb6cde5084f16feca

dp: implement XBUS DMA support

view details

Giovanni Bajo

commit sha 3575ce9ebfae0e33c1e2f0744b0be3a7dfe088dd

sp: add debug window with VU registers

view details

Giovanni Bajo

commit sha 6c182f452bb6c1574705299fc72763e431bf372b

Add workaround to IPF emulation bug to allow libdragon to boot

view details

Giovanni Bajo

commit sha 1d51ac544b47970338283faa5aa680d3f34619f9

mi: implement version register

view details

Giovanni Bajo

commit sha 8f5ea805b4777fce1ca8d51584db539109cf2bd6

emu: simplify DecodedInsn removing generic arguments

view details

Giovanni Bajo

commit sha 5cfda85a1ef60df16ac0b7c6110e0e3e8bdb9197

mips64: reuse existing constant register names

view details

push time in 3 months

issue commentocornut/imgui

Docking branch available for testing

@ocornut without any commitment, what is the ETA for this branch? Weeks? Months? Years? I’m starting to use it (through a binding) and I’d love to understand how much to engineer my build system through several layers of bindings and libraries to use this branch (I need several local forks/patches). I don’t want to waste much time on it if the branch is landing “soon”.

ocornut

comment created time in 3 months

issue openedrust-lang/rls

RLS freezes on "building" or "build_script_build"

RLS is stuck in the "building" or "build_script_build" phase. There is not generated output, save for this error any time I save a file:

  Message: Reformat failed to complete successfully
  Code: -32603 ```

I tried activate the log file and it's always zero bytes (nothing is written there).

How can I debug this issue? 

created time in 3 months

push eventrasky/imgui-rs

Malik Olivier Boussejra

commit sha 7268d35be72740e937a033bc3f81a922691c9096

[gfx_examples] Update glutin to 0.19.0 Attempt to fix #172 for gfx back-end.

view details

Malik Olivier Boussejra

commit sha d7a37a25fc3f1a8a82f1879df2c14f870050773b

[gfx-examples] Update Cargo.lock

view details

Malik Olivier Boussejra

commit sha 4ae0392e48738f92bb6a91cdfe5ec712756a2d9d

Bump minimum Rust version to 1.28 The previous commit added support for glutin 0.19.0. glutin 0.19.0 depends on winit 0.18.0, which requires rust 1.28.0 or later. Here is the full dependency chain: glutin 0.19.0 > winit 0.18 > smithay-client-toolkit 0.4.2 > andrew 0.1.4 > rusttype 0.7.3 > ordered-float 1.0.1 (on linux) ordered-float uses [repr(transparent)], which is now stabilised on latest rust, hence the requirement for rust 1.28.0.

view details

Antoine Gersant

commit sha 62e382a1db72512a7994d50941226f92ad3c26e3

Pulled master from Gekkio/imgui-rs

view details

Antoine Gersant

commit sha d61e69eee2ea68b88744ef3418594d2cfba99d01

Fixed cimgui version

view details

Joonas Javanainen

commit sha 44db8d81e67286090fd47da01bb770403b0ee5fe

Upgrade to imgui 1.66b

view details

Malik Olivier Boussejra

commit sha 0b459cf104f4e76b213fa9c1b95eb2646639048a

Merge pull request #183 from malikolivier/172-update-glutin-on-gfx Update glutin to 0.19.0 on gfx examples

view details

Antoine Gersant

commit sha 2b154986546309bc57ff374aba984243dd8590a7

Exposed a few functions related to content regions

view details

Antoine Gersant

commit sha cd2f7570e620cfb24e49a4cd3dd9c456ec6f9ba4

Added is_item_hovered_with_flags

view details

Antoine Gersant

commit sha 17b82e5cbb7a66d011ef7e6a778b0fa413f8ab78

Added is_window_hovered_with_flags

view details

Joonas Javanainen

commit sha eb5aaf1c2a758df8f63ab93949860e586c3df2c1

Replace imgui-glutin-support with imgui-winit-support

view details

Joonas Javanainen

commit sha 14eaf7720098f166853768d1984fa8a764e2c9e1

Merge pull request #189 from agersant/hover-flags Added ability to pass flags to is_item_hovered

view details

Joonas Javanainen

commit sha 490ffe056d14832f13e8e0e1e99451dad9cec13e

Merge pull request #185 from agersant/content-region Content region

view details

Joonas Javanainen

commit sha 7d364f6e5d3e3ce9a747208c788c53fdb60e88e6

Fix incorrect function name

view details

shockham

commit sha 5ace64bbe443b3ef2b6ca199f14a75c7457b3d6c

Update glium to 0.23

view details

Joonas Javanainen

commit sha 38bb92a8b90dd1d3b5b88c8fcaa7264f121f268e

Merge pull request #193 from shockham/master Update glium to 0.23

view details

Joonas Javanainen

commit sha 79829d320054a21240f084811e687e0c3f34d912

Bump minimum Rust version to 1.31 Damnit stb_truetype :'(

view details

Joonas Javanainen

commit sha 20139851e290de3f4d45c107fb9c4a3e71f1ab35

Upgrade to Rust 2018 edition We might as well do this since we've been forced to 1.31...

view details

Joonas Javanainen

commit sha 0d878e1ecbd6d006e16d8e13db7d184b54de6883

Upgrade dependencies in examples

view details

Joonas Javanainen

commit sha 91a961e06a11201947737b6d3e70a19350cb8829

Use more 2018 edition idioms

view details

push time in 3 months

PR opened Rust-SDL2/rust-sdl2

Upgrade to SDL 2.0.10

Updates both the bundled headers and build.rs for the bundled version. Tested on MacOS and Windows.

+760 -181

0 comment

88 changed files

pr created time in 3 months

push eventrasky/rust-sdl2

push time in 3 months

issue closedRust-SDL2/rust-sdl2

Error compiling master for unstable features

Compiling master, I get many errors like these:

error[E0658]: use of unstable library feature 'maybe_uninit' (see issue #53491)
    --> /Users/rasky/.cargo/git/checkouts/rust-sdl2-420a1e366198f510/7be0aee/src/sdl2/event.rs:1750:19
     |
1750 |     let mut raw = mem::MaybeUninit::uninit();
     |                   ^^^^^^^^^^^^^^^^^^^^^^^^
     |
     = help: add #![feature(maybe_uninit)] to the crate attributes to enable

error[E0658]: use of unstable library feature 'try_from' (see issue #33417)
 --> /Users/rasky/.cargo/git/checkouts/rust-sdl2-420a1e366198f510/7be0aee/src/sdl2/pixels.rs:3:5
  |
3 | use std::convert::TryFrom;
  |     ^^^^^^^^^^^^^^^^^^^^^
  |
  = help: add #![feature(try_from)] to the crate attributes to enable

I don't see those features being activated in lib.rs. Am I missing something?

closed time in 3 months

rasky

issue commentRust-SDL2/rust-sdl2

Error compiling master for unstable features

Updated toolchain, now it works.

rasky

comment created time in 3 months

push eventrasky/rust-sdl2

Giovanni Bajo

commit sha b5949d0933ae94199d0a140cda10a8057b571ab4

Activate unstable features

view details

push time in 3 months

issue openedRust-SDL2/rust-sdl2

Error compiling master for unstable features

Compiling master, I get many errors like these:

error[E0658]: use of unstable library feature 'maybe_uninit' (see issue #53491)
    --> /Users/rasky/.cargo/git/checkouts/rust-sdl2-420a1e366198f510/7be0aee/src/sdl2/event.rs:1750:19
     |
1750 |     let mut raw = mem::MaybeUninit::uninit();
     |                   ^^^^^^^^^^^^^^^^^^^^^^^^
     |
     = help: add #![feature(maybe_uninit)] to the crate attributes to enable

error[E0658]: use of unstable library feature 'try_from' (see issue #33417)
 --> /Users/rasky/.cargo/git/checkouts/rust-sdl2-420a1e366198f510/7be0aee/src/sdl2/pixels.rs:3:5
  |
3 | use std::convert::TryFrom;
  |     ^^^^^^^^^^^^^^^^^^^^^
  |
  = help: add #![feature(try_from)] to the crate attributes to enable

I don't see those features being activated in lib.rs. Am I missing something?

created time in 3 months

push eventrasky/rust-sdl2

Giovanni Bajo

commit sha 7be0aee7d65dba906d189732ead5fb5e3fab02ff

Upgrade to SDL 2.0.10

view details

push time in 3 months

push eventrasky/rust-sdl2

Michael Fairley

commit sha 23b26ed167693d112811601ecd753a5eea8b9b78

Mark Event as Send+Sync

view details

Sunjay Varma

commit sha 37cf5f76ac30bc19bd388cab64ae238bfd9983a3

Replaced all calls to unwrap() in examples with either the ? operator or expect

view details

Giovanni Bajo

commit sha 48a4a8e5c51faf2b8c7570b87366bbea60d435e3

Add missing shell32 to static link libraries on Windows

view details

Michael Fairley

commit sha ee8438a15e9e42b3c79e8fca2e3c5746c6bebce6

Add doc comments

view details

Cobrand

commit sha f5102401d572a9db026eb6d50aea62ad669b7719

Merge pull request #838 from michaelfairley/event_send_sync Mark Event as Send+Sync

view details

Hannes Steffenhagen

commit sha c1ce3a4d0fd9c8467464805a361a31fc942f0ceb

Use ::std instead of just std in rect.rs Without this, newer rustc versions complain with the error: > access to extern crates through prelude is experimental (see issue #44660)

view details

Sergey

commit sha 7058d783552b4da0e1822d6b59e374a0489448dc

Update README.md

view details

Seivan Heidari

commit sha a731b12f09015eab9486f7a8b0a90896f5d295a0

Homebrew packages are called libSDL2main and libSDL2 It can't find `SDL2main` or `SDL2` when trying to statically link with homebrew libs.

view details

Gregor Best

commit sha 31e879adb46c0f608fa38aca73e45b4e24b02091

Fix linking on OpenBSD

view details

german gomez

commit sha ff625a6e0d6cb5a436971a1d5b240765cce969ff

Window opacity Signed-off-by: german gomez <germangb42@gmail.com>

view details

Sebastian Verschoor

commit sha 94125a5dda082fd2f3d22d8e82ed8b0274204ae9

Added inplace operations for Point

view details

german gomez

commit sha f4cef9eea0f5e1d9618630c1d04235fbe8e998aa

Window opacity docs Signed-off-by: german gomez <germangb42@gmail.com>

view details

Cobrand

commit sha fb41194717469a6394b81e09293c9d10800e8583

Merge pull request #867 from germangb/missing_methods Window opacity methods

view details

germangb

commit sha 1d97e22ddbf1fca8b1f3e66c84cbafa86dbf9498

Bump minimum SDL version in README

view details

german gomez

commit sha d732cde2a137d757deb712908d784b8f9edf2cc3

Update changelog Signed-off-by: german gomez <germangb42@gmail.com>

view details

Sebastian Verschoor

commit sha 979d57d70af32af509c92643d9fdfd69dcdf4b9a

Updated changelog

view details

Sebastian Verschoor

commit sha e5de512bd4e19f3b0093621a26053487ec3b5e52

Updated changelog, (hopefully) without messing it up this time...

view details

Cobrand

commit sha 33c5710ac3da0c46d5b1e9c3f189e4da0b86af0b

Merge pull request #871 from germangb/minim_version Update minimum SDL2 version in README

view details

Cobrand

commit sha 0ebf34495fd21cb7e034a3fce29592abd4b73c94

Merge pull request #868 from sebastianv89/master Added inplace operations for Point

view details

Cobrand

commit sha 1d5f701d0f34d63eb4cb738d2f436e64e7ff3dae

Enable no_std for sdl2-sys

view details

push time in 3 months

issue commentGekkio/imgui-rs

SDL2 support.

@Gekkio any time I'm updating imgui-rs, I'm running into version dependency hell because I also need to update the above sdl2/opengl libraries. Any plan on merging them into imgui-rs?

thelearnerofcode

comment created time in 3 months

issue openedcimgui/cimgui

Failed generation on docking branch

Generation on docking branch fails:

struct ImGuiViewportP : public ImGuiViewport
{
    int Idx;
    int LastFrameActive;
    int LastFrameDrawLists[2];
    int LastFrontMostStampCount;
    ImGuiID LastNameHash;
    ImVec2 LastPos;
    float Alpha;
    float LastAlpha;
    short PlatformMonitor;
    bool PlatformWindowCreated;
    ImGuiWindow* Window;
    ImDrawList* DrawLists[2];
    ImDrawData DrawDataP;
    ImDrawDataBuilder DrawDataBuilder;
    ImVec2 LastPlatformPos;
    ImVec2 LastPlatformSize;
    ImVec2 LastRendererSize;
    ImGuiViewportP() { Idx = -1; LastFrameActive = LastFrameDrawLists[0] = LastFrameDrawLists[1] = LastFrontMostStampCount = -1; LastNameHash = 0; Alpha = LastAlpha = 1.0f; PlatformMonitor = -1; PlatformWindowCreated = false; Window = ((void *)0); DrawLists[0] = DrawLists[1] = ((void *)0); LastPlatformPos = LastPlatformSize = LastRendererSize = ImVec2(FLT_MAX, FLT_MAX); }
    ~ImGuiViewportP() { if (DrawLists[0]) IM_DELETE(DrawLists[0]); if (DrawLists[1]) IM_DELETE(DrawLists[1]); }
    ImRect GetRect() const { return ImRect(Pos.x, Pos.y, Pos.x + Size.x, Pos.y + Size.y); }
    ImVec2 GetCenter() const { return ImVec2(Pos.x + Size.x * 0.5f, Pos.y + Size.y * 0.5f); }
    void ClearRequestFlags() { PlatformRequestClose = PlatformRequestMove = PlatformRequestResize = false; }
};
luajit: ./cpp2ffi.lua:800: could not get stname
stack traceback:
	[C]: in function 'error'
	./cpp2ffi.lua:800: in function 'clean_struct'
	./cpp2ffi.lua:876: in function 'gen_structs_and_enums'
	./cpp2ffi.lua:713: in function 'do_parse'
	./generator.lua:575: in main chunk
	[C]: at 0x0109354300

created time in 3 months

issue closedalexcrichton/toml-rs

Serialize as hex

I have some specific integers I'd like to be serialized in hex format. It looks like this is currently not supported?

closed time in 3 months

rasky

issue commentalexcrichton/toml-rs

Serialize as hex

Found duplicate #285

rasky

comment created time in 3 months

issue openedalexcrichton/toml-rs

Serialize as hex

I have some specific integers I'd like to be serialized in hex format. It looks like this is currently not supported?

created time in 3 months

issue openedDragonMinded/libdragon

Support files > 16M (or at least abort when creating the filesystem)

DFS can encode only file sizes up to 2^24 because it uses a uint32_t flags that uses only 24 bits to encode file size, and 8 bit for flags:

https://github.com/DragonMinded/libdragon/blob/122a61546bac17f36537b1b73d2ea8758e188f3f/tools/mkdfs/mkdfs.c#L216

I think size and flags could safely be splitted into two different variables, considering that the same structure contains a pathname with MAX_PATH_LEN size, so a single additional byte for flags shouldn't matter much. The rest of the code (dfs_read, etc.) seems not to have problems with files larger than 16M.

Alternatively, if this is not wanted, at least mkdfs should abort when a file's size is bigger than 16M, rather than silently truncating it.

created time in 3 months

PR opened DataDog/zstd

Add a banner to mention the data corruption bug

Until https://github.com/DataDog/zstd/issues/22 is fixed, this library is using zstd in a way that can cause data corruption, as confirmed by zstd maintainer himself. I think this is critical enough that should be mentioned at the top of README and the Go community should be alerted until the bug is fixed.

+4 -0

0 comment

1 changed file

pr created time in 3 months

push eventrasky/zstd

Giovanni Bajo

commit sha 404b9bd215b63823358a4822ac78601c230aa0bb

Add a banner to mention the data corruption bug Until https://github.com/DataDog/zstd/issues/22 is fixed, this library is using zstd in a way that can cause data corruption, as confirmed by zstd maintainer himself. I think this is critical enough that should be mentioned at the top of README and the Go community should be alerted until the bug is fixed

view details

push time in 3 months

fork rasky/zstd

Zstd wrapper for Go

fork in 3 months

startedDragonMinded/libdragon

started time in 3 months

PR opened DragonMinded/libdragon

Upgrade binutils to fix RSP assembly alignment

Currently, people following the ucode/Makefile pattern of using objcopy to create sections of RSP binary code can hit an annoying bug in which the RSP symbol is not aligned to a 16-byte boundary, as required by load_ucode / load_data to operate correctly the DMA engine.

There are many possible workaround for this (including adding one of the many bin2h tools to the whole toolchain) but it looks like binutils just added a feature to objcopy. This is extracted from the 2.33 release notes:

  • Add --set-section-alignment <section-name>=<power-of-2-align> option to objcopy to allow the changing of section alignments.

So it seems that makefiles could be adjusted to use that option to force the RSP binary code to always be 16-byte aligned.

This PR upgrades binutils; I don't know how the official docker image is updated, but it would help a lot if it could get updated as well. Meanwhile, I'm also sending another PR to workaround the problem in load_ucode / load_data.

+1 -1

0 comment

1 changed file

pr created time in 3 months

push eventrasky/libdragon

Giovanni Bajo

commit sha 8a07f0b7c1df3fb214c61aa0bdc2360a02eaa199

Upgrade binutils to fix RSP assembly alignment Currently, people following the `ucode/Makefile` pattern of using `objcopy` to create sections of RSP binary code can hit an annoying bug in which the RSP symbol is not aligned to a 16-byte boundary, as required by `load_ucode` / `load_data` to operate correctly the DMA engine. There are many possible workaround for this (including adding one of the many `bin2h` tools to the whole toolchain) but it looks like binutils just added a feature to `objcopy`. This is extracted from the [2.33 release notes](https://sourceware.org/ml/binutils/2019-10/msg00103.html): > * Add --set-section-alignment <section-name>=<power-of-2-align> > option to objcopy to allow the changing of section alignments. So it seems that makefiles could be adjusted to use that option to force the RSP binary code to always be 16-byte aligned. This PR upgrades binutils; I don't know how the official docker image is updated, but it would help a lot if it could get updated as well. Meanwhile, I'm also sending another PR to workaround the problem in `load_ucode` / `load_data`.

view details

push time in 3 months

PR opened DragonMinded/libdragon

Fix wrong parentheses in CachedAddr

Parentheses were not balanced, leading to compilation errors.

+1 -1

0 comment

1 changed file

pr created time in 4 months

push eventrasky/libdragon

Giovanni Bajo

commit sha cf22bf30f7dba17018d534c654860891c6c84abc

Fix wrong parentheses in CachedAddr

view details

push time in 4 months

fork rasky/libdragon

Open source library for N64 development.

www.dragonminded.com/n64dev

fork in 4 months

issue commentgolang/go

runtime: defer is slow

Can we close this issue? It doesn't seem there's much left. Defers in a loop are always going to be somewhat slower.

minux

comment created time in 4 months

issue commentgolang/go

runtime: defer is slow

How’s that defer “in a loop”? It does look like it from a source code point of view, but in SSA it is in a block which is no more “part of the loop” than the first block after loop in source code order. It sounds like the loop detection made by the defer optimization is slightly off.

minux

comment created time in 4 months

issue commentgolang/go

cmd/compile: internal compiler error: aliasing constant which is not registered

The bug happens because of a logic snafu in the master's version poset.

poset.constants is a map from int64 constant values to poset indices, so if constants[5] is 32 it means that the 32th node of the poset contains the constant 5. This normally means that its correspondent SSA node is a constant.

When poset learns that two nodes are equal, it aliases them (which means that it collapses them to the same node, adjusting in/out edges). When collapsing two nodes one of each is a constant, either node is kept, and the other is removed (it doesn't really matter which). If the removed node is in the constant map, the constant map should be updated to point to the other node.

The bug happens because the aliasnode functions doesn't explicitly check whether the removed node is reference in the constant map to avoid a O(n) loop through it; instead, it checks whether the corresponding SSA node is a constant (isGenericConstInt()). This is not always the same, because the corresponding SSA node could have been previously aliased to a constant; so it would be found in the constant map, even if it's not a constant itself.

I have a pending CL 200860 that change aliasnode into a more generic aliasnodes; in doing so, it fixes this bug, by changing the isGenericConstInt() call to an explicit loop through the constant map. The CL has been already +2 pending a comment, and I'm going to submit it soon.

If @randall77 agrees, we can just keep this bug open one day or two until I submit that CL. We can then add the testcase to fixedbugs.

ALTree

comment created time in 4 months

more