Date: Fri, 06 Apr 2012 04:06:51 -0500 From: Alan Cox <alc@rice.edu> To: Konstantin Belousov <kostikbel@gmail.com> Cc: alc@freebsd.org, freebsd-hackers@freebsd.org, Andrey Zonov <andrey@zonov.org> Subject: Re: problems with mmap() and disk caching Message-ID: <4F7EB22B.3000509@rice.edu> In-Reply-To: <20120406083858.GG2358@deviant.kiev.zoral.com.ua> References: <4F7B495D.3010402@zonov.org> <20120404071746.GJ2358@deviant.kiev.zoral.com.ua> <4F7DC037.9060803@rice.edu> <20120405173138.GX2358@deviant.kiev.zoral.com.ua> <4F7DE3AD.5080401@rice.edu> <20120406083858.GG2358@deviant.kiev.zoral.com.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On 04/06/2012 03:38, Konstantin Belousov wrote: > On Thu, Apr 05, 2012 at 01:25:49PM -0500, Alan Cox wrote: >> On 04/05/2012 12:31, Konstantin Belousov wrote: >>> On Thu, Apr 05, 2012 at 10:54:31AM -0500, Alan Cox wrote: >>>> On 04/04/2012 02:17, Konstantin Belousov wrote: >>>>> On Tue, Apr 03, 2012 at 11:02:53PM +0400, Andrey Zonov wrote: >>>>>> Hi, >>>>>> >>>>>> I open the file, then call mmap() on the whole file and get pointer, >>>>>> then I work with this pointer. I expect that page should be only once >>>>>> touched to get it into the memory (disk cache?), but this doesn't work! >>>>>> >>>>>> I wrote the test (attached) and ran it for the 1G file generated from >>>>>> /dev/random, the result is the following: >>>>>> >>>>>> Prepare file: >>>>>> # swapoff -a >>>>>> # newfs /dev/ada0b >>>>>> # mount /dev/ada0b /mnt >>>>>> # dd if=/dev/random of=/mnt/random-1024 bs=1m count=1024 >>>>>> >>>>>> Purge cache: >>>>>> # umount /mnt >>>>>> # mount /dev/ada0b /mnt >>>>>> >>>>>> Run test: >>>>>> $ ./mmap /mnt/random-1024 30 >>>>>> mmap: 1 pass took: 7.431046 (none: 262112; res: 32; super: >>>>>> 0; other: 0) >>>>>> mmap: 2 pass took: 7.356670 (none: 261648; res: 496; super: >>>>>> 0; other: 0) >>>>>> mmap: 3 pass took: 7.307094 (none: 260521; res: 1623; super: >>>>>> 0; other: 0) >>>>>> mmap: 4 pass took: 7.350239 (none: 258904; res: 3240; super: >>>>>> 0; other: 0) >>>>>> mmap: 5 pass took: 7.392480 (none: 257286; res: 4858; super: >>>>>> 0; other: 0) >>>>>> mmap: 6 pass took: 7.292069 (none: 255584; res: 6560; super: >>>>>> 0; other: 0) >>>>>> mmap: 7 pass took: 7.048980 (none: 251142; res: 11002; super: >>>>>> 0; other: 0) >>>>>> mmap: 8 pass took: 6.899387 (none: 247584; res: 14560; super: >>>>>> 0; other: 0) >>>>>> mmap: 9 pass took: 7.190579 (none: 242992; res: 19152; super: >>>>>> 0; other: 0) >>>>>> mmap: 10 pass took: 6.915482 (none: 239308; res: 22836; super: >>>>>> 0; other: 0) >>>>>> mmap: 11 pass took: 6.565909 (none: 232835; res: 29309; super: >>>>>> 0; other: 0) >>>>>> mmap: 12 pass took: 6.423945 (none: 226160; res: 35984; super: >>>>>> 0; other: 0) >>>>>> mmap: 13 pass took: 6.315385 (none: 208555; res: 53589; super: >>>>>> 0; other: 0) >>>>>> mmap: 14 pass took: 6.760780 (none: 192805; res: 69339; super: >>>>>> 0; other: 0) >>>>>> mmap: 15 pass took: 5.721513 (none: 174497; res: 87647; super: >>>>>> 0; other: 0) >>>>>> mmap: 16 pass took: 5.004424 (none: 155938; res: 106206; super: >>>>>> 0; other: 0) >>>>>> mmap: 17 pass took: 4.224926 (none: 135639; res: 126505; super: >>>>>> 0; other: 0) >>>>>> mmap: 18 pass took: 3.749608 (none: 117952; res: 144192; super: >>>>>> 0; other: 0) >>>>>> mmap: 19 pass took: 3.398084 (none: 99066; res: 163078; super: >>>>>> 0; other: 0) >>>>>> mmap: 20 pass took: 3.029557 (none: 74994; res: 187150; super: >>>>>> 0; other: 0) >>>>>> mmap: 21 pass took: 2.379430 (none: 55231; res: 206913; super: >>>>>> 0; other: 0) >>>>>> mmap: 22 pass took: 2.046521 (none: 40786; res: 221358; super: >>>>>> 0; other: 0) >>>>>> mmap: 23 pass took: 1.152797 (none: 30311; res: 231833; super: >>>>>> 0; other: 0) >>>>>> mmap: 24 pass took: 0.972617 (none: 16196; res: 245948; super: >>>>>> 0; other: 0) >>>>>> mmap: 25 pass took: 0.577515 (none: 8286; res: 253858; super: >>>>>> 0; other: 0) >>>>>> mmap: 26 pass took: 0.380738 (none: 3712; res: 258432; super: >>>>>> 0; other: 0) >>>>>> mmap: 27 pass took: 0.253583 (none: 1193; res: 260951; super: >>>>>> 0; other: 0) >>>>>> mmap: 28 pass took: 0.157508 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 29 pass took: 0.156169 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 30 pass took: 0.156550 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> >>>>>> If I ran this: >>>>>> $ cat /mnt/random-1024> /dev/null >>>>>> before test, when result is the following: >>>>>> >>>>>> $ ./mmap /mnt/random-1024 5 >>>>>> mmap: 1 pass took: 0.337657 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 2 pass took: 0.186137 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 3 pass took: 0.186132 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 4 pass took: 0.186535 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> mmap: 5 pass took: 0.190353 (none: 0; res: 262144; super: >>>>>> 0; other: 0) >>>>>> >>>>>> This is what I expect. But why this doesn't work without reading file >>>>>> manually? >>>>> Issue seems to be in some change of the behaviour of the reserv or >>>>> phys allocator. I Cc:ed Alan. >>>> I'm pretty sure that the behavior here hasn't significantly changed in >>>> about twelve years. Otherwise, I agree with your analysis. >>>> >>>> On more than one occasion, I've been tempted to change: >>>> >>>> pmap_remove_all(mt); >>>> if (mt->dirty != 0) >>>> vm_page_deactivate(mt); >>>> else >>>> vm_page_cache(mt); >>>> >>>> to: >>>> >>>> vm_page_dontneed(mt); >>>> >>>> because I suspect that the current code does more harm than good. In >>>> theory, it saves activations of the page daemon. However, more often >>>> than not, I suspect that we are spending more on page reactivations than >>>> we are saving on page daemon activations. The sequential access >>>> detection heuristic is just too easily triggered. For example, I've >>>> seen it triggered by demand paging of the gcc text segment. Also, I >>>> think that pmap_remove_all() and especially vm_page_cache() are too >>>> severe for a detection heuristic that is so easily triggered. >>> Yes, I agree that such change shall be an improvement, and I expect >>> that Andrey will test it. >>> >>> On the other hand, I do think that allocator should prefer unnamed >>> pages to pages which still have valid content. On my 12G desktop, >>> I never saw more then 100MB of cached pages, and similar numbers >>> are observed on the 32-48GB servers. I suppose that this is related. >> On allocation, the system does prefer free pages over cached pages. >> When cached pages are added to the physical memory allocator, they are >> added to VM_FREEPOOL_CACHE. When pages are allocated, they are taken >> from VM_FREEPOOL_DEFAULT. Generally, pages only move from the CACHE >> pool to the DEFAULT pool when the DEFAULT pool is depleted. (However, >> occasionally, they do move because of coalescing.) When I redid the >> physical memory allocator, I looked at the rate of cached page >> reactivation under the old and the new allocators. At least for the >> tests that I did the rates weren't that different. It was low, >> single-digit percentages. I think the highest likelihood of >> reactivation comes from the pages that are cached by the sequential >> access heuristic because it is so overzealous. >> >> I don't think it's related. You see modest numbers of cached pages >> simply because the page daemon met its target for the sum of free and >> cached pages. So, it just stopped moving pages from the inactive queue >> into the physical memory allocator's cache/free queues. > No, I mean something else. Specifically, I mean that somehow the > preference for non-named pages does not work. At least, I cannot give > any other explanation for the following experiment. > > Lets take stock HEAD without change in vm_fault.c. The initial > state of 8GB machine is as follows, the test file was not even > stat(2)-ed yet. > Mem: 37M Active, 18M Inact, 150M Wired, 236K Cache, 27M Buf, 7612M Free > > Now, run the unmodified original Andrey' test with only one pass, > making sequential read of the mmap of a 5GB file from UFS volume. > After the run > Mem: 38M Active, 18M Inact, 153M Wired, 21M Cache, 30M Buf, 7586M Free > > Please note that cached count increased only for 20M, and this is > for calls to vm_page_cache() worth of 5GB. In other words, it seems > that allocator almost never touches free memory, always preferring > cache. This is mostly coincides with what I saw when I profiled > original problem reported by Andrey. Ah, I understand.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4F7EB22B.3000509>