Changes

353 bytes added ,  19:21, 31 July 2019
a bit specify what's structure of vdso
Line 5: Line 5:  
=== A problem ===
 
=== A problem ===
   −
While kernel guarantees backwards compatibility, i.e. helper functions won't suddenly disappear, the internal structure of the '''vDSO''' itself may vary the way it wants, userspace programs won't notice such changes but it brings a huge problem to CRIU. The '''vDSO''' structure is highly coupled with the kernel internals so that if application is migrating to a new kernel release with brand new '''vDSO''' old one (which is still sitting in application memory) strictly speaking is no longer usable, because the new kernel implies new '''vDSO''' internal structure.
+
While the kernel guarantees backwards compatibility (i.e. helper functions won't suddenly disappear), the internal structure of the '''vDSO''' itself may vary. Userspace programs won't notice such changes but it brings a huge problem to CRIU. The '''vDSO''' structure is highly coupled with the kernel internals. If an application is being migrated to a new kernel release (with a brand new '''vDSO''') the old '''vDSO''' (which is stored in the application's memory) is no longer usable.
    
=== Calls proxification ===
 
=== Calls proxification ===
Line 11: Line 11:  
To solve this problem CRIU does that named call proxification, where all functions from the original '''vDSO''' code are redirected to a new one provided by the kernel where application is restored. This done in a several steps on restore:
 
To solve this problem CRIU does that named call proxification, where all functions from the original '''vDSO''' code are redirected to a new one provided by the kernel where application is restored. This done in a several steps on restore:
   −
# Scan runtime environment for current '''vDSO''' and parse its structure
+
# Scan runtime environment for current '''vDSO''' and parse its structure (size, sections, symbols, etc)
 
# On restore of dumped '''vDSO''' memory area we patch function entry points so that they are redirected to current '''vDSO''' (for x86 architecture is simply a <code>jmp</code> instruction)
 
# On restore of dumped '''vDSO''' memory area we patch function entry points so that they are redirected to current '''vDSO''' (for x86 architecture is simply a <code>jmp</code> instruction)
   Line 17: Line 17:     
In case if an application is dumped and restored on same kernel CRIU detects that '''vDSO''' structure has not be changed and doesn't use calls proxification.
 
In case if an application is dumped and restored on same kernel CRIU detects that '''vDSO''' structure has not be changed and doesn't use calls proxification.
 +
 +
=== Unsolved proxification problems ===
 +
 +
# If an application in very unlikely case is Checkpointed over first bytes of vdso entry ''and'' vdso is different on the destination migration node, those might be the bytes that are being patched during proxification. If it's on vdso just after those entry-bytes, the stale data from vvar is taken (might be just fine afterall).
 +
 +
[[Category: Under the hood]]
105

edits