Skip to content
GitLab
Projects
Groups
Snippets
Help
Loading...
Help
What's new
10
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Open sidebar
Iris
Actris
Commits
f803fca1
Commit
f803fca1
authored
Nov 25, 2019
by
Robbert Krebbers
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Tweak README.
parent
2773194c
Pipeline
#21578
passed with stage
in 5 minutes and 20 seconds
Changes
1
Pipelines
1
Hide whitespace changes
Inline
Sidebyside
Showing
1 changed file
with
22 additions
and
33 deletions
+22
33
README.md
README.md
+22
33
No files found.
README.md
View file @
f803fca1
...
...
@@ 16,14 +16,15 @@ In order to build, install the above dependencies and then run
## Theory of Actris
The theory of Actris (semantics of channels, the
CPS
model, and the proof rules)
The theory of Actris (semantics of channels, the model, and the proof rules)
can be found in the directory
[
theories/channel
](
theories/channel
)
. The files
correspond to the following parts of the paper:

[
theories/channel/channel.v
](
theories/channel/channel.v
)
: The definitional
semantics of bidirectional channels in terms of Iris's HeapLang language.

[
theories/channel/proto_model.v
](
theories/channel/proto_model.v
)
: The CPS
model of Dependent Separation Protocols over arbitrary BIlogics.

[
theories/channel/proto_model.v
](
theories/channel/proto_model.v
)
: The
construction of the model of Dependent Separation Protocols as the solution of
a recursive domain equation.

[
theories/channel/proto_channel.v
](
theories/channel/proto_channel.v
)
: The
instantiation of protocols with the Iris logic, definition of the connective
`↣`
for channel endpoint ownership, and lemmas corresponding to the Actris proof rules.
...
...
@@ 44,28 +45,18 @@ correspond to the following parts of the paper:
## Notation
The notation for Dependent Separation Protocols differ between the
mechani
s
ation
and the paper.
The notation for Dependent Separation Protocols differ between the
paper and
the Coq
mechani
z
ation
:
The paper uses the following notation:

Send: ! x_1 .. x_n
<
v
>
{ P } . prot

Recv: ? x_1 .. x_n
<
v
>
{ P } . prot

End: end

Select: prot_1 {Q_1} ⊕ {Q_2} prot_2

Branch: prot_1 {Q_1} & {Q_2} prot_2

Append: prot_1 · prot_2

Dual: An overlined protocol
The mechanisation uses the following notation:

Send:
<
!
>
x_1 .. x_n, MSG v {{ P }} ; prot

Recv:
<
?
>
x_1 .. x_n, MSG v {{ P }} ; prot

End: END

Select: prot_1
<
{
Q_1
}
+
{
Q_2
}
>
prot_2

Branch: prot_1
<
{
Q_1
}
&
{
Q_2
}
>
prot_2

Append: prot_1
<
++
>
prot_2

Dual: Nothing
  Paper  Coq mechanization 

 Send 
`! x_1 .. x_n <v>{ P }. prot`

`<!> x_1 .. x_n, MSG v {{ P }}; prot`

 Recv 
`? x_1 .. x_n <v>{ P }. prot`

`<?> x_1 .. x_n, MSG v {{ P }}; prot`

 End 
`end`

`END`

 Select 
`prot_1 {Q_1}⊕{Q_2} prot_2`

`prot_1 <{Q_1}+{Q_2}> prot_2`

 Branch 
`prot_1 {Q_1}&{Q_2} prot_2`

`prot_1 <{Q_1}&{Q_2}> prot_2`

 Append 
`prot_1 · prot_2`

`prot_1 <++> prot_2`

 Dual  An overlined protocol  No notation 
## Weakest preconditions and Coq tactics
...
...
@@ 188,13 +179,11 @@ of Actris and the formalization in Coq, that are briefly discussed here.
2.
the higherorder ghost state used for ownership of protocols, and
3.
the opening of the protocol invariant.

**Subtyping relation in c↣ prot**

**Protocol subtyping**
The mechanization has introduced the notion of "protocol subtyping", which
allows one to strengthen/weaken the predicates of sends/receives, respectively.
This achieved using the relation
`iProto_le p p'`
, and the additional rule
`c ↣ p ∗ iProto_le p p' ∗ c ↣ p'`
. To support "protocol subtyping", the
definition of
`c ↣ p`
in the model is changed to be closed under
`iProto_le`
.
The mechanisation has introduced the notion of subtyping, which allows one to
strengten/weaken the predicates of sends/receives respectively. In particular
this means that the endpoint ownership has been extended with
`iProto_le p p'`
,
where
`p'`
is the original protocol used in the ghost state, and
`p`
is the
protocol denoted by the ownership
`c↣ prot`
. The effect of this is that the
user can update his own view of the protocol, as long as it is consistent
with the original protocol in the invariant. As such the fundamental aspect of
the ownership still align with that of the paper.
\ No newline at end of file
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment