Hi, regard to Gem5, you *MUST *use the same disk image that you have created the checkpoint from. In order to accelerate the file transferring and perform execution, the best way (I use it) is to run Gem5 with *two* disk images. The second one, you can use it for moving files (during simulating session). Assume this scenario ... start Gem5 with two disk images (main_image and temp_image). While the Gem5 is running, we can perform the following ... mount the temp_image and copy all binaries/benchmarks we want, umount it. At this point, the Gem5 still running correctly without any problem ... now from inside gem5 terminal ... mount the second disk image, i.e., temp_image ... and you will find your files that we recently moved them. ... You can perform this task frequently, faster ... and ang changes to the second disk image, i.e., temp_image, will remain permanently.
The following command worked for me ... - ./build/ARM/gem5.opt - --dot-config=Det_L2.dot configs/example/fs.py - --checkpoint-dir=m5out/ -r 1 --restore-with-cpu=DerivO3CPU - *--disk-image=/home/null/Desktop/ARM-img/Up-2-date/IMG/2/edited_img/aarch64.img * - *--disk-image=/home/null/Desktop/ARM-img/Up-2-date/IMG/2/edited_img/spare.img* - --kernel=/home/null/Desktop/ARM-img/2018/binaries/vmlinux.vexpress_gem5_v1_64 - --cpu-type=DerivO3CPU - --caches --l2cache --l1d_size=64kB --l1i_size=64kB --l2_size=1024kB - --mem-size=4GB Now ... where the second disk image is? from the terminal's Gem5 ... you should be able to find the second disk image in /dev/sdb1 So ... you need (like working on host) to mount it to /mnt and you will find your files. Use this [Create spare-empty disk image <https://yulistic.gitlab.io/2016/05/gem5-make-a-new-boot-image/>] to create an empty disk image. On Tue, Jan 12, 2021 at 1:49 AM Sherif AbdelFadil via gem5-users < [email protected]> wrote: > Hello, > I am trying to figure out a way to modify my disk image and move large > files onto it as fast as possible, and then use that image for a full > system simulation. > I tried mounting the disk image as a loopback device then copying an > executable onto it. When I use the disk image with qemu, everything works > fine: I can find the executable and run it. However, when I restore a gem5 > FS simulation from a checkpoint using the modified image, I either can't > find the added files, or sometimes I get an error message that they/the > file system are/is corrupted. When I go back to qemu, everything is still > fine. > Any ideas on what the issue is and how I can fix it? > _______________________________________________ > gem5-users mailing list -- [email protected] > To unsubscribe send an email to [email protected] > %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
_______________________________________________ gem5-users mailing list -- [email protected] To unsubscribe send an email to [email protected] %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
