Page MenuHomeFreeBSD
Differential D30408

tests/libalias: Test LibAliasIn with redirection
ClosedPublic
Actions

Authored by donner on May 23 2021, 8:54 PM.
Tags
None
Referenced Files
Unknown Object (File)
Sat, Apr 12, 1:42 PM
Unknown Object (File)
Thu, Apr 10, 5:28 AM
Unknown Object (File)
Tue, Apr 1, 3:00 AM
Unknown Object (File)
Wed, Mar 26, 9:59 AM
Unknown Object (File)
Sun, Mar 23, 12:32 AM
Unknown Object (File)
Mar 14 2025, 9:27 AM
Unknown Object (File)
Mar 12 2025, 9:55 AM
Unknown Object (File)
Mar 11 2025, 2:19 AM
View All Files
Subscribers
asomers
imp
melifaro

Details

Reviewers
kp
lev
Group Reviewers
network
Commits
rG755bab6d55e6: tests/libalias: Test LibAliasIn on redirection
Summary

Extend the test suite for libalias(3) to incoming connections.
Test the various types of redirections.

Test Plan

make test

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Passed
Unit
No Test Coverage
Build Status
Buildable 39374
Build 36263: arc lint + arc unit

Event Timeline

donner created this revision.May 23 2021, 8:54 PM
Herald added subscribers: melifaro, asomers, imp. · View Herald TranscriptMay 23 2021, 8:54 PM
donner requested review of this revision.May 23 2021, 8:54 PM
Harbormaster completed remote builds in B39364: Diff 89699.May 23 2021, 8:54 PM
donner added a reviewer: network.May 23 2021, 8:56 PM

Factor out the common test infrastructure and start with the first test.

donner updated this revision to Diff 89700.May 23 2021, 9:34 PM
  • Clear state instead of choosing new flows
Harbormaster completed remote builds in B39365: Diff 89700.May 23 2021, 9:34 PM
donner updated this revision to Diff 89701.May 23 2021, 9:42 PM
  • Add final variant of PortRedirect
Harbormaster completed remote builds in B39366: Diff 89701.May 23 2021, 9:42 PM
donner updated this revision to Diff 89703.May 23 2021, 9:55 PM
  • test overlapping portforwardings
Harbormaster completed remote builds in B39368: Diff 89703.May 23 2021, 9:55 PM
donner updated this revision to Diff 89718.May 24 2021, 1:00 PM
  • Clear state instead of choosing new flows
  • Add final variant of PortRedirect
  • test overlapping portforwardings
  • Refactor UDP tests

Rebase to D30412

Harbormaster completed remote builds in B39374: Diff 89718.May 24 2021, 1:00 PM
donner added a parent revision: D30412: tests/libalias: Factor out common primitives.May 24 2021, 1:01 PM
donner updated this revision to Diff 89720.May 24 2021, 1:03 PM

Depend on D30412

Harbormaster completed remote builds in B39376: Diff 89720.May 24 2021, 1:03 PM
donner updated this revision to Diff 89721.May 24 2021, 1:38 PM
  • Add test for change of alias address
Harbormaster completed remote builds in B39377: Diff 89721.May 24 2021, 1:38 PM
donner updated this revision to Diff 89723.May 24 2021, 1:47 PM
  • Apply missing refactors
Harbormaster completed remote builds in B39378: Diff 89723.May 24 2021, 1:47 PM
donner updated this revision to Diff 89724.May 24 2021, 2:11 PM
  • Add tests for LibAliasRedirectAddr
Harbormaster completed remote builds in B39379: Diff 89724.May 24 2021, 2:11 PM
donner updated this revision to Diff 89725.May 24 2021, 2:28 PM
  • Add test for LSNAT
Harbormaster completed remote builds in B39380: Diff 89725.May 24 2021, 2:28 PM
donner updated this revision to Diff 89728.May 24 2021, 3:36 PM
  • Add test for LibAliasRedirectDynamic
Harbormaster completed remote builds in B39381: Diff 89728.May 24 2021, 3:37 PM
donner updated this revision to Diff 89775.May 24 2021, 8:11 PM

Rebase

Harbormaster completed remote builds in B39413: Diff 89775.May 24 2021, 8:11 PM
donner updated this revision to Diff 89776.May 24 2021, 8:12 PM
  • Add copyright statement
Harbormaster completed remote builds in B39414: Diff 89776.May 24 2021, 8:12 PM
donner mentioned this in D30277: libalias: tidy up housekeeping.May 25 2021, 7:22 AM
This revision was not accepted when it landed; it landed in state Needs Review.May 26 2021, 2:23 PM
Closed by commit rG755bab6d55e6: tests/libalias: Test LibAliasIn on redirection (authored by donner). · Explain Why
This revision was automatically updated to reflect the committed changes.
donner added a commit: rG755bab6d55e6: tests/libalias: Test LibAliasIn on redirection.
donner mentioned this in rG488f39f355ab: tests/libalias: Test LibAliasIn and redirection.Jun 2 2021, 9:36 AM
donner mentioned this in rG19888502bf39: tests/libalias: Test LibAliasIn and redirection.Jun 2 2021, 9:38 AM

Revision Contents
Changeset List

PathSize
tests/
sys/
netinet/
libalias/
245 lines
2 lines
1 line

Diff 89718

View Options

tests/sys/netinet/libalias/3_natin.c

Show All 40 Lines
* *
* CRC32 code derived from work by Gary S. Brown. * CRC32 code derived from work by Gary S. Brown.
*/ */
#include <sys/cdefs.h> #include <sys/cdefs.h>
__FBSDID("$FreeBSD$"); __FBSDID("$FreeBSD$");
#include <sys/param.h> #include <sys/param.h>
#include <sys/libkern.h>
#include <sys/systm.h> #include <sys/systm.h>
#if defined(__amd64__) || defined(__i386__)
#include <machine/md_var.h>
#include <machine/specialreg.h>
#endif
const uint32_t crc32_tab[] = { const uint32_t crc32_tab[] = {
0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f, 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f,
0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988,
0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2,
0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7,
0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9, 0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172, 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172,
0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c,
▲ Show 20 Lines • Show All 685 Lines • ▼ Show 20 Linesmultitable_crc32c(uint32_t crc32c,
return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word)); return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word));
} }
uint32_t uint32_t
calculate_crc32c(uint32_t crc32c, calculate_crc32c(uint32_t crc32c,
const unsigned char *buffer, const unsigned char *buffer,
unsigned int length) unsigned int length)
{ {
if (length < 4) { if (length < 4) {
jhbUnsubmitted
Not Done
Inline Actions

You can just use #ifdef __SSE4_2__. -msse42 (or -msse4) will define that via the compiler driver. That is what controls 'popcnt' (in the case of popcnt we use it if you set CPUTYPE in your make.conf to specify a CPU that includes popcnt).

jhb: You can just use `#ifdef __SSE4_2__`. -msse42 (or -msse4) will define that via the compiler…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is -msse4 used to compile this file on i386/amd64? I thought it was off by default, and only enabled on a few files. My build logs actually show -mno-sse -mno-avx.

cem: Is `-msse4` used to compile this file on i386/amd64? I thought it was off by default, and only…
markjUnsubmitted
Not Done
Inline Actions

The entries in sys/conf/files.* explicitly specify -msse4.

markj: The entries in sys/conf/files.* explicitly specify -msse4.
cemAuthorUnsubmitted
Not Done
Inline Actions

Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).

cem: Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).
return (singletable_crc32c(crc32c, buffer, length)); return (singletable_crc32c(crc32c, buffer, length));
#if defined(__amd64__) || defined(__i386__)
kibUnsubmitted
Not Done
Inline Actions

Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way out) ?

kib: Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is used outside of the kernel.

PC98 is excluded because I believe it predates SSE4.2, and I preferred to avoid including dead SSE4.2 code in the PC98 kernel if it would never be used.

cem: Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is…
} else if ((cpu_feature2 & CPUID2_SSE42) != 0) {
return (sse42_crc32c(crc32c, buffer, length));
#endif
} else { } else {
return (multitable_crc32c(crc32c, buffer, length)); return (multitable_crc32c(crc32c, buffer, length));
} }
} }
View Options

tests/sys/netinet/libalias/Makefile

Show All 40 Lines
* *
* CRC32 code derived from work by Gary S. Brown. * CRC32 code derived from work by Gary S. Brown.
*/ */
#include <sys/cdefs.h> #include <sys/cdefs.h>
__FBSDID("$FreeBSD$"); __FBSDID("$FreeBSD$");
#include <sys/param.h> #include <sys/param.h>
#include <sys/libkern.h>
#include <sys/systm.h> #include <sys/systm.h>
#if defined(__amd64__) || defined(__i386__)
#include <machine/md_var.h>
#include <machine/specialreg.h>
#endif
const uint32_t crc32_tab[] = { const uint32_t crc32_tab[] = {
0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f, 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f,
0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988,
0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2,
0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7,
0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9, 0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172, 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172,
0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c,
▲ Show 20 Lines • Show All 685 Lines • ▼ Show 20 Linesmultitable_crc32c(uint32_t crc32c,
return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word)); return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word));
} }
uint32_t uint32_t
calculate_crc32c(uint32_t crc32c, calculate_crc32c(uint32_t crc32c,
const unsigned char *buffer, const unsigned char *buffer,
unsigned int length) unsigned int length)
{ {
if (length < 4) { if (length < 4) {
jhbUnsubmitted
Not Done
Inline Actions

You can just use #ifdef __SSE4_2__. -msse42 (or -msse4) will define that via the compiler driver. That is what controls 'popcnt' (in the case of popcnt we use it if you set CPUTYPE in your make.conf to specify a CPU that includes popcnt).

jhb: You can just use `#ifdef __SSE4_2__`. -msse42 (or -msse4) will define that via the compiler…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is -msse4 used to compile this file on i386/amd64? I thought it was off by default, and only enabled on a few files. My build logs actually show -mno-sse -mno-avx.

cem: Is `-msse4` used to compile this file on i386/amd64? I thought it was off by default, and only…
markjUnsubmitted
Not Done
Inline Actions

The entries in sys/conf/files.* explicitly specify -msse4.

markj: The entries in sys/conf/files.* explicitly specify -msse4.
cemAuthorUnsubmitted
Not Done
Inline Actions

Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).

cem: Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).
return (singletable_crc32c(crc32c, buffer, length)); return (singletable_crc32c(crc32c, buffer, length));
#if defined(__amd64__) || defined(__i386__)
kibUnsubmitted
Not Done
Inline Actions

Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way out) ?

kib: Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is used outside of the kernel.

PC98 is excluded because I believe it predates SSE4.2, and I preferred to avoid including dead SSE4.2 code in the PC98 kernel if it would never be used.

cem: Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is…
} else if ((cpu_feature2 & CPUID2_SSE42) != 0) {
return (sse42_crc32c(crc32c, buffer, length));
#endif
} else { } else {
return (multitable_crc32c(crc32c, buffer, length)); return (multitable_crc32c(crc32c, buffer, length));
} }
} }
View Options

tests/sys/netinet/libalias/util.h

Show All 40 Lines
* *
* CRC32 code derived from work by Gary S. Brown. * CRC32 code derived from work by Gary S. Brown.
*/ */
#include <sys/cdefs.h> #include <sys/cdefs.h>
__FBSDID("$FreeBSD$"); __FBSDID("$FreeBSD$");
#include <sys/param.h> #include <sys/param.h>
#include <sys/libkern.h>
#include <sys/systm.h> #include <sys/systm.h>
#if defined(__amd64__) || defined(__i386__)
#include <machine/md_var.h>
#include <machine/specialreg.h>
#endif
const uint32_t crc32_tab[] = { const uint32_t crc32_tab[] = {
0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f, 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f,
0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988,
0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2,
0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7,
0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9, 0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172, 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172,
0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c,
▲ Show 20 Lines • Show All 685 Lines • ▼ Show 20 Linesmultitable_crc32c(uint32_t crc32c,
return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word)); return (crc32c_sb8_64_bit(crc32c, buffer, length, to_even_word));
} }
uint32_t uint32_t
calculate_crc32c(uint32_t crc32c, calculate_crc32c(uint32_t crc32c,
const unsigned char *buffer, const unsigned char *buffer,
unsigned int length) unsigned int length)
{ {
if (length < 4) { if (length < 4) {
jhbUnsubmitted
Not Done
Inline Actions

You can just use #ifdef __SSE4_2__. -msse42 (or -msse4) will define that via the compiler driver. That is what controls 'popcnt' (in the case of popcnt we use it if you set CPUTYPE in your make.conf to specify a CPU that includes popcnt).

jhb: You can just use `#ifdef __SSE4_2__`. -msse42 (or -msse4) will define that via the compiler…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is -msse4 used to compile this file on i386/amd64? I thought it was off by default, and only enabled on a few files. My build logs actually show -mno-sse -mno-avx.

cem: Is `-msse4` used to compile this file on i386/amd64? I thought it was off by default, and only…
markjUnsubmitted
Not Done
Inline Actions

The entries in sys/conf/files.* explicitly specify -msse4.

markj: The entries in sys/conf/files.* explicitly specify -msse4.
cemAuthorUnsubmitted
Not Done
Inline Actions

Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).

cem: Only for libkern/x86/crc32_sse42.c. Not libkern/crc32.c (this file).
return (singletable_crc32c(crc32c, buffer, length)); return (singletable_crc32c(crc32c, buffer, length));
#if defined(__amd64__) || defined(__i386__)
kibUnsubmitted
Not Done
Inline Actions

Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way out) ?

kib: Do we need defined(_KERNEL) there as well ? Why PC98 is excluded (it seems to be on the way…
cemAuthorUnsubmitted
Not Done
Inline Actions

Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is used outside of the kernel.

PC98 is excluded because I believe it predates SSE4.2, and I preferred to avoid including dead SSE4.2 code in the PC98 kernel if it would never be used.

cem: Is this code built outside of _KERNEL? tinderbox succeeds, so I don't believe this code is…
} else if ((cpu_feature2 & CPUID2_SSE42) != 0) {
return (sse42_crc32c(crc32c, buffer, length));
#endif
} else { } else {
return (multitable_crc32c(crc32c, buffer, length)); return (multitable_crc32c(crc32c, buffer, length));
} }
} }