swh:1:snp:a72e953ecd624a7df6e6196bbdd05851996c5e40
Tip revision: bf534986350a991e4a1b29126de0342ffd76205e authored by Alex Arslan on 06 February 2022, 15:21:33 UTC
[deps] Update PCRE2 URL (#43884) (#44047)
[deps] Update PCRE2 URL (#43884) (#44047)
Tip revision: bf53498
secretbuffer.jl
# This file is a part of Julia. License is MIT: https://julialang.org/license
"""
Base.SecretBuffer()
An [`IOBuffer`](@ref)-like object where the contents will be securely wiped when garbage collected.
It is considered best practice to wipe the buffer using `Base.shred!(::SecretBuffer)` as
soon as the secure data are no longer required. When initializing with existing data, the
`SecretBuffer!` method is highly recommended to securely zero the passed argument. Avoid
initializing with and converting to `String`s as they are unable to be securely zeroed.
# Examples
```jldoctest
julia> s = Base.SecretBuffer()
SecretBuffer("*******")
julia> write(s, 's', 'e', 'c', 'r', 'e', 't')
6
julia> seek(s, 0); Char(read(s, UInt8))
's': ASCII/Unicode U+0073 (category Ll: Letter, lowercase)
julia> Base.shred!(s)
SecretBuffer("*******")
julia> eof(s)
true
```
"""
mutable struct SecretBuffer <: IO
data::Vector{UInt8}
size::Int
ptr::Int
function SecretBuffer(; sizehint=128)
s = new(Vector{UInt8}(undef, sizehint), 0, 1)
finalizer(final_shred!, s)
return s
end
end
"""
SecretBuffer(str::AbstractString)
A convenience constructor to initialize a `SecretBuffer` from a non-secret string.
Strings are bad at keeping secrets because they are unable to be securely
zeroed or destroyed. Therefore, avoid using this constructor with secret data.
Instead of starting with a string, either construct the `SecretBuffer`
incrementally with `SecretBuffer()` and [`write`](@ref), or use a `Vector{UInt8}` with
the `Base.SecretBuffer!(::Vector{UInt8})` constructor.
"""
SecretBuffer(str::AbstractString) = SecretBuffer(String(str))
function SecretBuffer(str::String)
buf = codeunits(str)
s = SecretBuffer(sizehint=length(buf))
for c in buf
write(s, c)
end
seek(s, 0)
s
end
convert(::Type{SecretBuffer}, s::AbstractString) = SecretBuffer(String(s))
"""
SecretBuffer!(data::Vector{UInt8})
Initialize a new `SecretBuffer` from `data`, securely zeroing `data` afterwards.
"""
function SecretBuffer!(d::Vector{UInt8})
len = length(d)
s = SecretBuffer(sizehint=len)
for i in 1:len
write(s, d[i])
end
seek(s, 0)
securezero!(d)
s
end
unsafe_SecretBuffer!(s::Cstring) = unsafe_SecretBuffer!(convert(Ptr{UInt8}, s), Int(ccall(:strlen, Csize_t, (Cstring,), s)))
function unsafe_SecretBuffer!(p::Ptr{UInt8}, len=1)
s = SecretBuffer(sizehint=len)
for i in 1:len
write(s, unsafe_load(p, i))
end
seek(s, 0)
unsafe_securezero!(p, len)
s
end
show(io::IO, s::SecretBuffer) = print(io, "SecretBuffer(\"*******\")")
# Unlike other IO objects, equality is computed by value for convenience
==(s1::SecretBuffer, s2::SecretBuffer) = (s1.ptr == s2.ptr) && (s1.size == s2.size) && (UInt8(0) == _bufcmp(s1.data, s2.data, min(s1.size, s2.size)))
# Also attempt a constant time buffer comparison algorithm — the length of the secret might be
# inferred by a timing attack, but not its values.
@noinline function _bufcmp(data1::Vector{UInt8}, data2::Vector{UInt8}, sz::Int)
res = UInt8(0)
for i = 1:sz
res |= xor(data1[i], data2[i])
end
return res
end
# All SecretBuffers hash the same to avoid leaking information or breaking consistency with ==
const _sb_hash = UInt === UInt32 ? 0x111c0925 : 0xb06061e370557428
hash(s::SecretBuffer, h::UInt) = hash(_sb_hash, h)
function write(io::SecretBuffer, b::UInt8)
if io.ptr > length(io.data)
# We need to resize! the array: do this manually to ensure no copies are left behind
newdata = Vector{UInt8}(undef, (io.size+16)*2)
copyto!(newdata, io.data)
securezero!(io.data)
io.data = newdata
end
io.size == io.ptr-1 && (io.size += 1)
io.data[io.ptr] = b
io.ptr += 1
return 1
end
function write(io::IO, s::SecretBuffer)
nb = 0
for i in 1:s.size
nb += write(io, s.data[i])
end
return nb
end
cconvert(::Type{Cstring}, s::SecretBuffer) = unsafe_convert(Cstring, s)
function unsafe_convert(::Type{Cstring}, s::SecretBuffer)
# Ensure that no nuls appear in the valid region
if any(==(0x00), s.data[i] for i in 1:s.size)
throw(ArgumentError("`SecretBuffers` containing nul bytes cannot be converted to C strings"))
end
# Add a hidden nul byte just past the end of the valid region
p = s.ptr
s.ptr = s.size + 1
write(s, '\0')
s.ptr = p
s.size -= 1
return Cstring(unsafe_convert(Ptr{Cchar}, s.data))
end
seek(io::SecretBuffer, n::Integer) = (io.ptr = max(min(n+1, io.size+1), 1); io)
seekend(io::SecretBuffer) = seek(io, io.size+1)
skip(io::SecretBuffer, n::Integer) = seek(io, position(io) + n)
bytesavailable(io::SecretBuffer) = io.size - io.ptr + 1
position(io::SecretBuffer) = io.ptr-1
eof(io::SecretBuffer) = io.ptr > io.size
isempty(io::SecretBuffer) = io.size == 0
function peek(io::SecretBuffer, ::Type{UInt8})
eof(io) && throw(EOFError())
return io.data[io.ptr]
end
function read(io::SecretBuffer, ::Type{UInt8})
eof(io) && throw(EOFError())
byte = io.data[io.ptr]
io.ptr += 1
return byte
end
function final_shred!(s::SecretBuffer)
!isshredded(s) && @async @warn("a SecretBuffer was `shred!`ed by the GC; use `shred!` manually after use to minimize exposure.")
shred!(s)
end
function shred!(s::SecretBuffer)
securezero!(s.data)
s.ptr = 1
s.size = 0
return s
end
isshredded(s::SecretBuffer) = all(iszero, s.data)
function shred!(f::Function, x)
try
f(x)
finally
shred!(x)
end
end