Kernel: Scheduler Woes, Buffered I/O and Userspace Pages

-
The Linux Kernel's Scheduler Apparently Causing Issues For Google Stadia Game Developers
Among the issues that game developers have been facing in bringing their games to Linux for Google's Stadia cloud gaming service apparently stem from kernel scheduler issues. We've known the Linux kernel scheduler could use some improvements and independent developers like Con Kolivas with BFS / MuQSS have pushed for such, but hopefully in 2020 we'll see some real action.
Game/C++ developer Malte Skarupke wrote a post about how bad the Linux kernel scheduler is and that solutions like MuQSS are an improvement but not complete. Malte noted, "I found that most mutex implementations are really good, that most spinlock implementations are pretty bad, and that the Linux scheduler is OK but far from ideal. The most popular replacement, the MuQSS scheduler has other problems instead. (the Windows scheduler is pretty good though)."
-
Buffered I/O without page-cache thrashing
Linux offers two modes for file I/O: buffered and direct. Buffered I/O passes through the kernel's page cache; it is relatively easy to use and can yield significant performance benefits for data that is accessed multiple times. Direct I/O, instead, goes straight between a user-space buffer and the storage device. It can be much faster for situations where caching by the operating system isn't necessary, but it is complex to use and contains traps for the unwary. Now, it seems, Jens Axboe has come up with a way to get many of the benefits of direct I/O with a lot less bother.
Direct I/O can give better performance than buffered I/O in a couple of ways. One of those is simply avoiding the cost of copying the data between user space and the page cache; that cost can be significant, but in many cases it is not the biggest problem. The real issue may be the effect of buffered I/O on the page cache.A process that performs large amounts of buffered I/O spread out over one or more large (relative to available memory) files will quickly fill the page cache (and thus memory) with cached file data. If the process in question does not access those pages after performing I/O, there is no benefit to keeping the data in memory, but it's there anyway. To be able to allocate memory for other uses, the kernel will have to reclaim some pages from somewhere. That can be expensive for the system as a whole, even if "somewhere" is the data associated with this I/O activity.
-
Explicit pinning of user-space pages
The saga of get_user_pages() — and the problems it causes within the kernel — has been extensively chronicled here; see the LWN kernel index for the full series. In short, get_user_pages() is used to pin user-space pages in memory for some sort of manipulation outside of the owning process(es); that manipulation can sometimes surprise other parts of the kernel that think they have exclusive rights to the pages in question. This patch series from John Hubbard does not solve all of the problems, but it does create some infrastructure that may make a solution easier to come by.
To simplify the situation somewhat, the problems with get_user_pages() come about in two ways. One of those happens when the kernel thinks that the contents of a page will not change, but some peripheral device writes new data there. The other arises with memory that is located on persistent-memory devices managed by a filesystem; pinning pages into memory deprives the filesystem of the ability to make layout changes involving those pages. The latter problem has been "solved" for now by disallowing long-lasting page pins on persistent-memory devices, but there are use cases calling for creating just that kind of pin, so better solutions are being sought.
Part of the problem comes down to the fact that get_user_pages() does not perform any sort of special tracking of the pages it pins into RAM. It does increment the reference count for each page, preventing it from being evicted from memory, but pages that have been pinned in this way are indistinguishable from pages that have acquired references in any of a vast number of other ways. So, while one can ask whether a page has references, it is not possible for kernel code to ask whether a page has been pinned for purposes like DMA I/O.
-

- Login or register to post comments
Printer-friendly version- 2455 reads
PDF version
More in Tux Machines
- Highlights
- Front Page
- Latest Headlines
- Archive
- Recent comments
- All-Time Popular Stories
- Hot Topics
- New Members
digiKam 7.7.0 is released
After three months of active maintenance and another bug triage, the digiKam team is proud to present version 7.7.0 of its open source digital photo manager. See below the list of most important features coming with this release.
|
Dilution and Misuse of the "Linux" Brand
|
Samsung, Red Hat to Work on Linux Drivers for Future Tech
The metaverse is expected to uproot system design as we know it, and Samsung is one of many hardware vendors re-imagining data center infrastructure in preparation for a parallel 3D world.
Samsung is working on new memory technologies that provide faster bandwidth inside hardware for data to travel between CPUs, storage and other computing resources. The company also announced it was partnering with Red Hat to ensure these technologies have Linux compatibility.
|
today's howtos
|








.svg_.png)
Content (where original) is available under CC-BY-SA, copyrighted by original author/s.

Recent comments
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago
1 year 11 weeks ago