Don't force MBEDTLS_PLATFORM_SNPRINTF_ALT on Windows in check_config.h #3453
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
check_config.h forces MBEDTLS_PLATFORM_SNPRINTF_ALT and
MBEDTLS_PLATFORM_VSNPRINTF_ALT to be defined on Windows. The reason
for this is that the native (v)snprintf on older Windows runtimes (up
to Visual Studio 2015: _MSVC_VER <= 1900) does not correctly
zero-terminate the buffer. This is documented in platform.h and tested
in selftest.c. To work around this bug, the library defines wrappers
mbedtls_platform_win32_snprintf and mbedtls_platform_win32_vsnprintf
in platform.c and uses them as the value of mbedtls_snprintf and
mbedtls_vsnprintf.
Even if MBEDTLS_PLATFORM_{V,}SNPRINTF_ALT is not defined,
the macro mbedtls_{v,}snprintf expands to the corresponding wrapper
function on platforms where it is necessary. Therefore modifying the
configuration to set the xxxSNPRINTF_ALT symbols is not actually
necessary. Don't do it.
With this change, check_config.h is truly a configuration _check_, as
the name implies, and does not modify the configuration.
Move the requirement for MBEDTLS_PLATFORM_C on Windows platforms to
platform.h, and limit it to platforms where the snprintf wrapper is
necessary, since this is the only reason why the platform module is
necessary.
Signed-off-by: Gilles Peskine <[email protected]>
4 tasks
|
The CI jobs Windows-2013 and Windows-mingw pass but win32-mingw, win32-msvc12_64 and win32_msvc12_32 complain about a broken snprintf. Back to the drawing board. |
|
I am closing this pull request because it isn't working, I don't have much expertise on the topic (Windows) and I have no time to investigate further. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
check_config.hforcesMBEDTLS_PLATFORM_SNPRINTF_ALTandMBEDTLS_PLATFORM_VSNPRINTF_ALTto be defined on Windows. The reason for this is that the native (v)snprintf on older Windows runtimes (up to Visual Studio 2015:_MSVC_VER <= 1900) does not correctly zero-terminate the buffer. This is documented inplatform.hand tested inselftest.c. To work around this bug, the library defines wrappers mbedtls_platform_win32_snprintfandmbedtls_platform_win32_vsnprintfinplatform.cand uses them as the value ofmbedtls_snprintfandmbedtls_vsnprintf`.Even if
MBEDTLS_PLATFORM_{V,}SNPRINTF_ALTis not defined, the macrombedtls_{v,}snprintfexpands to the corresponding wrapper function on platforms where it is necessary. Therefore modifying the configuration to set thexxxSNPRINTF_ALTsymbols is not actually necessary. Don't do it.With this change,
check_config.his truly a configuration check, as the name implies, and does not modify the configuration.Here's a CI run with an additional patch that removes the wrapper altogether, to validate that the wrapper is getting used in this pull request: https://jenkins-internal.mbed.com/job/mbedtls-release-new/740/
snprintf(onlyvsnprintf). Never mind, the failure of snprintf on mingw shows that we do have a test that would catch the absence of the wrapper if it was necessary.I don't see a need to backport this. The only reason I want to make this change is as a precursor to changing the way
check_config.his included, as done in #1999. That would be too disruptive for LTS branches.