VeraCrypt 1.24 is faulty - Details Provided

[Mike Sanders]

This appears to be an issue others have reported as well.
Bob Stewart's post - 2020-01-01 "Can't create large hidden container on Debian Stretch"

There is a definite bug in 1.24 with creating hidden containers. After two weeks of systematic testing the following was done to narrow and identify the issue,

OS Tested -
linuxmint-19.3-cinnamon-64bit
CentOS-8-x86_64-1905-dvd1

Hardware Used - 4 Systems
Dell Precision & a XPS Series (intel) laptop
Samsung Series 3 Notebook (intel)
ASUS ROG (intel) laptop

Bios Types
Secure Boot w/ EUFI
EUFI
Legacy

Drives
SSD -- Traditional Drives -- USB drives
Samsung -- Western Digital -- Kingston, Sandisk, Samsung

OS was loaded on a independent drive/device

Steps
Create Volume
Create a volume with partition/drive
Hidden VeraCrypt volume
Select target device (see below)
Use Defaults (AES - SHA-512)
Type Password
Randomize and Format
Place files in outer format (200 files ~ 800 MB)
Next
Use Defaults (AES - SHA-512)
Failure at this point with verison 1.24 -- Success with 1.23

*not the drive or any partition where OS is installed, root of a secondary device like USB drive

After reading countless forums on the subject of usage of Veracrypt the following was ruled out.

OS - Issue recreated on both linux platforms Mint and CentOS
Hardware - That should be pretty obvious given the array of choices and configuration settings I must have loaded the OS a dozen items between hardware, and worn down my usb and ssd drives formating them.

What was consistent VeraCrypt 1.24 and 1.24 Update 3 (did not try other versions of 1.24) failed to allow one to use the total of the reported maximum size for a hidden container. Version 1.23 worked just fine on 3 of the systems with either OS in mixed bios modes.

It is of my opinon that version 1.24 suffers from a flaw that prohibits one from using the maximum possible size fora hidden volume. While the sizes vary between mediums of what was tested, the end result was a large reduction in what was availale to be used, despite the software indicating otherwise. (see images)

512GB drive had 5.8 GB max, any selection higher than 5800 would result in the next button being greyed out. Similiar results for USB 64GB device would report 57GB available but only allow 3400MB. The screen shots provided were for a 8GB USB stick that showed 7.4 GIB available but only 1930MB can be choosen, anything greater than 1930 would result in the next button being greyed out prohibiting you to move foward in the process. (see images)

This happened on multiple hardware configurations and on two different distributions of linux. Simply put, VeraCrypt 1.24 is faulty.

[Bob Stewart]

Thanks for verifying my results. That's quite a lot of work you must have done. I must have tried to create a 1.8TB hidden partition 8 or 10 times before I thought to go back to 1.23 and had success.

[Mike Sanders]

I havent used VeraCrypt before, and only lightly with TrueCrypt. The concept of a hidden volume was intresting. As some others have pointed out wear and tear make such a thing apparent. Therefore it is not so hidden.

Note: I understand the layout of plausible denyability, and wont argue the merits of that one way or the other.. different post/subject. Just making note that is not truely hidden bit of smoke and mirrors if you would.

However, I was quite annoyed when the process did not seem to work as intended. More annoyed having wasted the writes on various ssd compoents, wondering what the hell was I doing wrong. Had I missed something, further confounded about the layering of partitions inside one another. After reading lots and lots of posts. Some dating back when the gui/cli was different and not applicable to the 1.24 release. This only exacerbated the attempt to solve and understand the problem. On the flip side I know alot more how it works.

So off I set to determine what it wasnt, a rather arduous task as some of the systems are not particuarlly powerful. Submitting a post with the details will hopefully allow others to recreate the issue and determine the source of the issue. The problem can definately be recreated, so up to the developers to look into it when they can.

All things being said kudos to the staff having maintained suport for this important tool.

[Nick]

I never got a responce for a similar post here.
https://sourceforge.net/p/veracrypt/discussion/technical/thread/0cce515546/
Even on Windows, this is the first version I've had an issue with, ever. VeraCrypt 1.24 update 2, will not let you mount volumes from the main mounting window. Also, none of the links in the help menu or the homepage link works. It's like something is broke with the interface. I downgraded back to VeraCrypt 1.23 hotfix 1, and everyting works again.

[Jane]

I also have the same issue. Hope this can be resolved soon.

[Mounir IDRASSI]

@msanders, @quakeboy02: Indeed, this is a regression but there is a temporary a workaround by using the switch --no-size-check: if it is specified, the faulty code will not be executed and the issue will be fixed.

I will implement a permanent fix and probably publish a 1.24-Update4 version.

[Bob Stewart]

Thanks a lot. Looking forward to the updated version.