Details
-
Bug
-
Resolution: Unresolved
-
Critical
-
None
-
0.4.15-dev some changes within MM??? Or a leaking lwIP driver???
Description
Hooray to the new Memory Manager again. 2GB spent to VM, ran out of RAM nevertheless.
And that wouldn't have happened with releases/0.4.14
Tested here:
ReactOS-0.4.15-4-x86.iso GCC8.4.0 built with RosBE-Linux2.2.1 dbg x86.
I 'just' wanted to retry the reproduction steps for https://jira.reactos.org/browse/CORE-18516 (a scenario that worked very well in all releases/0.4.10 to releases/0.4.14), but that regressed during 0.4.15-dev'-MM-rewrite.
However, when I retried today, it failed even earlier during the git clone.
Reproduction steps
- create VBox 4.3.12 VM with at least 2000MB of RAM spent to VM (you will need that much!)
- the virtual HDD should be at least 10GB (you will need that much)
- install ros into the VM, I used "ReactOS-0.4.15-4-x86.iso GCC8.4.0 built with RosBE-Linux2.2.1 dbg x86" here, I installed on FAT
- install git 2.10.0 from rapps
- install RosBE 2.1.6 from rapps, use C:\ros as the source folder
- switch desktop resolution to 1152x864 to have enough space for convenient working
- run RosBE2.1.6
- Run the taskmgr in the performance tab, and leave it running to observe resource consumption. Place it next to RosBE2.1.6 window
- now within the RosBE2.1.6 prompt do:
- 'prompt $T $P$G'
- 'git config --system http.sslverify false'
- 'git clone https://github.com/reactos/reactos.git'
Expected result
it should return to the prompt after having cloned the sources
Observed result
it does not. Instead it failed during the clone with OutOfRAM
It always happens reproduibly after the output "100% compressing objects", approximately 3minutes afte the cloning was started
Ros invoked the RAM OutOfmemory-dumper :
(ntoskrnl/mm/ARM3/expool.c:507) Out of memory dumper!
|
There is still plenty of disk space free by the way
ReactOS-0.4.15-4-x86_ENOSPC.log
ReactOS-0.4.15-4-x86_ENOSPC1.png (the error on screen)
ReactOS-0.4.15-4-x86_ENOSPC2.png (shows free space on disk)
ReactOS-0.4.15-4-x86_ENOSPC3.png (taskmgr performance tab)
The Memory manager on 0.4.15 is still worse than the Memory Manager in 0.4.14
Attachments
Issue Links
- blocks
-
CORE-18516 Regression, building ros on ros RosBE2.1.6 'ninja bootcd -j1' does fail since some 0.4.15-dev commit at step237 during compiling boot/environ with random errors
- Open