From northmirko at gmail.com Tue Oct 19 13:59:07 2021 From: northmirko at gmail.com (Mirko Kovacevic) Date: Tue, 19 Oct 2021 15:59:07 +0200 Subject: TTCN-3 tests for MME Message-ID: Hi, Osmonians, I don't know where to write for MME TTCN-3, so I chose to write to it closet relative - SGSN. I was trying to do slight modification to TTCN-3 code, modifing S1AP_Templates.ttcn. I embedded enodeb name for all tests with new protoc-IE = id-eNBname : ################################## template (value) S1AP_PDU ts_S1AP_SetupReq(template (value) Global_ENB_ID p_global_ENB_ID, template (value) SupportedTAs p_supportedTAs, template (value) PagingDRX p_pagingDRXs) := { initiatingMessage := { procedureCode := id_S1Setup, criticality := reject, value_ := { S1SetupRequest := { protocolIEs := { { id := S1AP_Constants.id_Global_ENB_ID, criticality := ignore, value_ := { Global_ENB_ID := p_global_ENB_ID } }, { id := S1AP_Constants.id_SupportedTAs, criticality := reject, value_ := {SupportedTAs := p_supportedTAs} }, { id := S1AP_Constants.id_eNBname,//id-eNBname criticality := ignore, value_ := {ENBname := "OsmoEnodeb"} } /* HACK: work around nextepc bug , { id := S1AP_Constants.id_pagingDRX, criticality := ignore, value_ := {PagingDRX := p_pagingDRXs} } */ } } } } } ######################## Code compiles, but PDU is not encoded well. I attached Wireshark preview. Can you help me with this issue ? -- Puno pozdrava, Mirko Kovacevic -------------- next part -------------- A non-text attachment was scrubbed... Name: image.png Type: image/png Size: 13567 bytes Desc: not available URL: From northmirko at gmail.com Fri Oct 22 12:12:03 2021 From: northmirko at gmail.com (Mirko Kovacevic) Date: Fri, 22 Oct 2021 14:12:03 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: Hi, Protocol-IEs, MMEname and ENBname, cant be decoded\encoded properly, definitely. I am using open5g-mme. When IE=MMEname is turned off, test can pass or used till the end. Amazing progress with S1AP emulation, congrats. If I misuse this group with the topic, be free to tell me. On Tue, Oct 19, 2021 at 3:59 PM Mirko Kovacevic wrote: > Hi, Osmonians, > I don't know where to write for MME TTCN-3, so I chose to write to it > closet relative - SGSN. > > I was trying to do slight modification to TTCN-3 code, modifing > S1AP_Templates.ttcn. > > I embedded enodeb name for all tests with new protoc-IE = id-eNBname : > ################################## > template (value) S1AP_PDU > ts_S1AP_SetupReq(template (value) Global_ENB_ID p_global_ENB_ID, > template (value) SupportedTAs p_supportedTAs, > template (value) PagingDRX p_pagingDRXs) := { > initiatingMessage := { > procedureCode := id_S1Setup, > criticality := reject, > value_ := { > S1SetupRequest := { > protocolIEs := { > { > id := S1AP_Constants.id_Global_ENB_ID, > criticality := ignore, > value_ := { Global_ENB_ID := p_global_ENB_ID } > }, > { > id := S1AP_Constants.id_SupportedTAs, > criticality := reject, > value_ := {SupportedTAs := p_supportedTAs} > }, > { > id := S1AP_Constants.id_eNBname,//id-eNBname > criticality := ignore, > value_ := {ENBname := "OsmoEnodeb"} > } /* HACK: work around nextepc bug > , { > id := S1AP_Constants.id_pagingDRX, > criticality := ignore, > value_ := {PagingDRX := p_pagingDRXs} > } */ > } > } > } > } > } > ######################## > > Code compiles, but PDU is not encoded well. > I attached Wireshark preview. > > Can you help me with this issue ? > > > > -- > Puno pozdrava, > Mirko Kovacevic > -- *Puno pozdrava,Mirko Kovacevic* -------------- next part -------------- An HTML attachment was scrubbed... URL: From laforge at osmocom.org Mon Oct 25 15:46:40 2021 From: laforge at osmocom.org (Harald Welte) Date: Mon, 25 Oct 2021 17:46:40 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: On Fri, Oct 22, 2021 at 02:12:03PM +0200, Mirko Kovacevic wrote: > Protocol-IEs, MMEname and ENBname, cant be decoded\encoded properly, > definitely. one would have to check if the encoding problem already exists in the BER version (as generated by TITAN natively) or if it happens at the BER <-> PER transcoding inside the [unfortunately] non-public libfftranscode, generated by ffasn1c. I you can build a small, self contained test case that shows a problem in BER <-> PER transcoding using the libfftranscode API directly (removing all of TITAN, etc.) then we may have a chance of either solving it at sysmocom or by asking the ffasn1c author to have a look > Amazing progress with S1AP emulation, congrats. thanks. Please keep us posted about any progress. We're happy to merge any related patches you may have for adding more test cases, fixing bugs, etc. I never found the time for it, but the general idea always was to automatically run this testsuite in jenkins.osmocom.org against the open5gs-mme, like we do for or own osmocom software. Regards, Harald -- - Harald Welte http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) From laforge at osmocom.org Mon Oct 25 15:42:26 2021 From: laforge at osmocom.org (Harald Welte) Date: Mon, 25 Oct 2021 17:42:26 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: Hi Mirko, sorry for the late response.. On Tue, Oct 19, 2021 at 03:59:07PM +0200, Mirko Kovacevic wrote: > I don't know where to write for MME TTCN-3, so I chose to write to it > closet relative - SGSN. That's ok, we don't really have any mailing lists specific for the test suites. > Code compiles, but PDU is not encoded well. > I attached Wireshark preview. > > Can you help me with this issue ? I don't have the time to dive into this at the moment, sorry. It's been years since I last touched that code, and There are tons of other projects and issues requiring urgent attention :/ I did have some problems when first implementing it, see https://bugs.eclipse.org/bugs/show_bug.cgi?id=549252 -- - Harald Welte http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) From kyeongun15 at gmail.com Wed Oct 6 03:58:03 2021 From: kyeongun15 at gmail.com (Gyeongun Kang) Date: Wed, 06 Oct 2021 03:58:03 -0000 Subject: [PATCH net-next] gtp: use skb_dst_update_pmtu_no_confirm() instead of direct call Message-ID: <20211006035739.5377-1-kyeongun15@gmail.com> skb_dst_update_pmtu_no_confirm() is a just wrapper function of ->update_pmtu(). So, it doesn't change logic Signed-off-by: Gyeongun Kang --- drivers/net/gtp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 30e0a10595a1..24e5c54d06c1 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -539,7 +539,7 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, mtu = dst_mtu(&rt->dst); } - rt->dst.ops->update_pmtu(&rt->dst, NULL, skb, mtu, false); + skb_dst_update_pmtu_no_confirm(skb, mtu); if (!skb_is_gso(skb) && (iph->frag_off & htons(IP_DF)) && mtu < ntohs(iph->tot_len)) { -- 2.25.1 From patchwork-bot+netdevbpf at kernel.org Wed Oct 6 14:26:36 2021 From: patchwork-bot+netdevbpf at kernel.org (patchwork-bot+netdevbpf at kernel.org) Date: Wed, 06 Oct 2021 14:26:36 -0000 Subject: [PATCH net-next] gtp: use skb_dst_update_pmtu_no_confirm() instead of direct call In-Reply-To: <20211006035739.5377-1-kyeongun15@gmail.com> References: <20211006035739.5377-1-kyeongun15@gmail.com> Message-ID: <163353000797.15249.17539810847769225450.git-patchwork-notify@kernel.org> Hello: This patch was applied to netdev/net-next.git (refs/heads/master): On Wed, 6 Oct 2021 03:57:39 +0000 you wrote: > skb_dst_update_pmtu_no_confirm() is a just wrapper function of > ->update_pmtu(). So, it doesn't change logic > > Signed-off-by: Gyeongun Kang > --- > drivers/net/gtp.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) Here is the summary with links: - [net-next] gtp: use skb_dst_update_pmtu_no_confirm() instead of direct call https://git.kernel.org/netdev/net-next/c/5b71131b795f You are awesome, thank you! -- Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html From laforge at osmocom.org Wed Oct 27 18:33:03 2021 From: laforge at osmocom.org (Harald Welte) Date: Wed, 27 Oct 2021 20:33:03 +0200 Subject: Osmocom redmine upgrade / Let me know if something broke Message-ID: Dear Osmocom community, today we finally upgraded our redmine installation from the unmaintained 3.4.x to the latest 4.2.3. The upgrade had been overdue for years, but today we (actually, Kevin) finally managed to find out how to make the openid_provider plugin to work with modern rails 5.x. In any case, I'm just informing you in case there is some unexpected fallout. I've verified that at least the following appears working: * logging into redmine * openid authentication from gerrit.osmocom.org (also after logout/relogin) * graphviz rendering * mscgen rendering There could however still be unexpected problems, particularly in features such as * e-mail notifications from redmine * updating redmine issues via e-mail * git integration ("Closes: OS#xxxx") If you run into any troubles, pleas report to https://osmocom.org/projects/osmocom-servers or if that also fails, feel free to send e-mails. Regards, Harald -- - Harald Welte http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) From northmirko at gmail.com Wed Oct 27 20:43:42 2021 From: northmirko at gmail.com (Mirko Kovacevic) Date: Wed, 27 Oct 2021 22:43:42 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: Thank you for your kind suggestions. I made a simple test with code snippet (lines): 1. var S1AP_PDU ber_pdu_to_log := valueof(ts_S1AP_SetupReq(g_enb_pars[idx].global_enb_id,g_enb_pars[idx].supported_tas, v32)); 2. log(ber_pdu_to_log); 3. var charstring pdu_to_send_to_ip_layer := enc_S1AP_PDU(ber_pdu_to_log); 4. log(pdu_to_send_to_ip_layer); RESULT from line 2. { initiatingMessage := { procedureCode := 17, criticality := reject (0), value_ := { s1SetupRequest := { protocolIEs := { { id := 59, criticality := ignore (1), value_ := { global_ENB_ID := { pLMNidentity := '62F224'O, eNB_ID := { macroENB_ID := '00000001010110110011'B }, iE_Extensions := omit } } }, { id := 60, criticality := ignore (1), value_ := { eNBname := "Ksenija" } }, { id := 64, criticality := reject (0), value_ := { supportedTAs := { { tAC := '3039'O ("09"), broadcastPLMNs := { '62F224'O }, iE_Extensions := omit } } } } } } } } } ################################# RESULT from line 3. ASN.1 (A)PER S1AP '00110020000003003B40080062F22400015B30003C4002030000400007000C0E4062F224'O ################################# I checked encoded ASN.1 message with publicly available decoders. Error is thrown whenever decoder tries to decode ENBname. I can conclude that libfftranscode API is the source of the problem. On Mon, Oct 25, 2021 at 5:50 PM Harald Welte wrote: > > On Fri, Oct 22, 2021 at 02:12:03PM +0200, Mirko Kovacevic wrote: > > Protocol-IEs, MMEname and ENBname, cant be decoded\encoded properly, > > definitely. > > one would have to check if the encoding problem already exists in the BER version > (as generated by TITAN natively) or if it happens at the BER <-> PER transcoding > inside the [unfortunately] non-public libfftranscode, generated by ffasn1c. > > I you can build a small, self contained test case that shows a problem in > BER <-> PER transcoding using the libfftranscode API directly (removing all of TITAN, etc.) > then we may have a chance of either solving it at sysmocom or by asking > the ffasn1c author to have a look > > > Amazing progress with S1AP emulation, congrats. > > thanks. Please keep us posted about any progress. We're happy to merge any > related patches you may have for adding more test cases, fixing bugs, etc. > > I never found the time for it, but the general idea always was to automatically > run this testsuite in jenkins.osmocom.org against the open5gs-mme, like we do > for or own osmocom software. > > Regards, > Harald > -- > - Harald Welte http://laforge.gnumonks.org/ > ============================================================================ > "Privacy in residential applications is a desirable marketing option." > (ETSI EN 300 175-7 Ch. A6) -- Puno pozdrava, Mirko Kovacevic From northmirko at gmail.com Fri Oct 29 07:13:25 2021 From: northmirko at gmail.com (Mirko Kovacevic) Date: Fri, 29 Oct 2021 09:13:25 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: But don't bother with this small defect, we can live without MMEname, ENBname, they are optional in many production systems, too. S1 handover would be a very good case to write. On Wed, Oct 27, 2021 at 10:43 PM Mirko Kovacevic wrote: > Thank you for your kind suggestions. > I made a simple test with code snippet (lines): > > 1. var S1AP_PDU ber_pdu_to_log := > > valueof(ts_S1AP_SetupReq(g_enb_pars[idx].global_enb_id,g_enb_pars[idx].supported_tas, > v32)); > 2. log(ber_pdu_to_log); > 3. var charstring pdu_to_send_to_ip_layer := > enc_S1AP_PDU(ber_pdu_to_log); > 4. log(pdu_to_send_to_ip_layer); > > > > RESULT from line 2. > { initiatingMessage := { procedureCode := 17, criticality := reject > (0), value_ := { s1SetupRequest := { protocolIEs := { { id := 59, > criticality := ignore (1), value_ := { global_ENB_ID := { pLMNidentity > := '62F224'O, eNB_ID := { macroENB_ID := '00000001010110110011'B }, > iE_Extensions := omit } } }, { id := 60, criticality := ignore (1), > value_ := { eNBname := "Ksenija" } }, { id := 64, criticality := > reject (0), value_ := { supportedTAs := { { tAC := '3039'O ("09"), > broadcastPLMNs := { '62F224'O }, iE_Extensions := omit } } } } } } } } > } > ################################# > RESULT from line 3. ASN.1 (A)PER S1AP > '00110020000003003B40080062F22400015B30003C4002030000400007000C0E4062F224'O > ################################# > > > I checked encoded ASN.1 message with publicly available decoders. > Error is thrown whenever decoder tries to decode ENBname. > I can conclude that libfftranscode API is the source of the problem. > > On Mon, Oct 25, 2021 at 5:50 PM Harald Welte wrote: > > > > On Fri, Oct 22, 2021 at 02:12:03PM +0200, Mirko Kovacevic wrote: > > > Protocol-IEs, MMEname and ENBname, cant be decoded\encoded properly, > > > definitely. > > > > one would have to check if the encoding problem already exists in the > BER version > > (as generated by TITAN natively) or if it happens at the BER <-> PER > transcoding > > inside the [unfortunately] non-public libfftranscode, generated by > ffasn1c. > > > > I you can build a small, self contained test case that shows a problem in > > BER <-> PER transcoding using the libfftranscode API directly (removing > all of TITAN, etc.) > > then we may have a chance of either solving it at sysmocom or by asking > > the ffasn1c author to have a look > > > > > Amazing progress with S1AP emulation, congrats. > > > > thanks. Please keep us posted about any progress. We're happy to merge > any > > related patches you may have for adding more test cases, fixing bugs, > etc. > > > > I never found the time for it, but the general idea always was to > automatically > > run this testsuite in jenkins.osmocom.org against the open5gs-mme, like > we do > > for or own osmocom software. > > > > Regards, > > Harald > > -- > > - Harald Welte > http://laforge.gnumonks.org/ > > > ============================================================================ > > "Privacy in residential applications is a desirable marketing option." > > (ETSI EN 300 175-7 Ch. > A6) > > > > -- > Puno pozdrava, > Mirko Kovacevic > -- *Puno pozdrava,Mirko Kovacevic* -------------- next part -------------- An HTML attachment was scrubbed... URL: From laforge at gnumonks.org Fri Oct 29 11:00:46 2021 From: laforge at gnumonks.org (Harald Welte) Date: Fri, 29 Oct 2021 13:00:46 +0200 Subject: TTCN-3 tests for MME In-Reply-To: References: Message-ID: Hi Mirko, On Fri, Oct 29, 2021 at 09:13:25AM +0200, Mirko Kovacevic wrote: > But don't bother with this small defect, we can live without MMEname, > ENBname, they are optional in many production systems, too. Thanks. Given the current workload and the fact that there is no customer paying for this, I think it will be a low priority item and unlikely to be investigated any time soon. > S1 handover would be a very good case to write. Well, osmocom is a collaborative open source project. Anyone can add whatever test cases they want, we're happy to merge any related patches. Regards, Harald -- - Harald Welte http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)