RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-05-18 21:25:07 +00:00
Code Issues Activity Actions

Kernel: Use KBuffers for ProcFS and SynthFS

Browse source 
Instead of generating ByteBuffers and keeping those lying around, have
these filesystems generate KBuffers instead. These are way less spooky
to leave around for a while.

Since FileDescription will keep a generated file buffer around until
userspace has read the whole thing, this prevents trivially exhausting
the kmalloc heap by opening many files in /proc for example.

The code responsible for generating each /proc file is not perfectly
efficient and many of them still use ByteBuffers internally but they
at least go away when we return now. :^)
This commit is contained in:
Andreas Kling 2019-08-05 11:35:49 +02:00
parent 6b6b86fbf3
commit 79e22acb22
7 changed files with 58 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

14
Kernel/FileSystem/SyntheticFileSystem.cpp
View file

@ -50,7 +50,7 @@ NonnullRefPtr<SynthFSInode> SynthFS::create_text_file(String&& name, ByteBuffer&
auto file = adopt(*new SynthFSInode(*this, generate_inode_index()));
file->m_data = contents;
file->m_name = move(name);
file->m_metadata.size = file->m_data.size();
file->m_metadata.size = file->m_data.value().size();
file->m_metadata.uid = 100;
file->m_metadata.gid = 200;
file->m_metadata.mode = mode;
@ -194,7 +194,7 @@ ssize_t SynthFSInode::read_bytes(off_t offset, ssize_t count, u8* buffer, FileDe
ASSERT(offset >= 0);
ASSERT(buffer);
ByteBuffer generated_data;
Optional<KBuffer> generated_data;
if (m_generator) {
if (!description) {
generated_data = m_generator(const_cast<SynthFSInode&>(*this));
@ -205,9 +205,15 @@ ssize_t SynthFSInode::read_bytes(off_t offset, ssize_t count, u8* buffer, FileDe
}
}
auto* data = generated_data ? &generated_data : &m_data;
const KBuffer* data_to_use = nullptr;
if (generated_data.has_value())
data_to_use = &generated_data.value();
else if (m_data.has_value())
data_to_use = &m_data.value();
else
ASSERT_NOT_REACHED();
ssize_t nread = min(static_cast<off_t>(data->size() - offset), static_cast<off_t>(count));
memcpy(buffer, data->pointer() + offset, nread);
memcpy(buffer, data->data() + offset, nread);
if (nread == 0 && description && description->generator_cache())
description->generator_cache().clear();
return nread;