diff for duplicates of <20110412180933.GA2836@joana> diff --git a/a/1.txt b/N1/1.txt index 1ba4503..c16a844 100644 --- a/a/1.txt +++ b/N1/1.txt @@ -1,93 +1,79 @@ * Gustavo F. Padovan <padovan@profusion.mobi> [2011-04-11 19:25:04 -0300]: > * Thomas Gleixner <tglx@linutronix.de> [2011-04-12 00:19:32 +0200]: ->=20 +> > > On Tue, 12 Apr 2011, Thomas Gleixner wrote: > > > On Mon, 11 Apr 2011, Marcel Holtmann wrote: -> > >=20 +> > > > > > > Hi Thomas, -> > > >=20 -> > > > > > > > Can the bluetooth folks please have a look at that ASAP? Th= -e obvious -> > > > > > > > fast fix for Linus tree is to revert the second hunk for no= -w, but this +> > > > +> > > > > > > > Can the bluetooth folks please have a look at that ASAP? The obvious +> > > > > > > > fast fix for Linus tree is to revert the second hunk for now, but this > > > > > > > > needs to be fixed proper. -> > > > > > >=20 -> > > > > > > Who will submit this patch? I'd rather have your name on it s= -o that +> > > > > > > +> > > > > > > Who will submit this patch? I'd rather have your name on it so that > > > > > > > people come complain at you... -> > > > > >=20 -> > > > > > I took a shot at it and just sent a patch (also attached for co= -nvenience)=20 +> > > > > > +> > > > > > I took a shot at it and just sent a patch (also attached for convenience) > > > > > > that should solve the problem. -> > > > >=20 +> > > > > > > > > > Aaarg. No. That patch reverts both hunks. -> > > > >=20 +> > > > > > > > > > --- a/net/bluetooth/hci_core.c > > > > > +++ b/net/bluetooth/hci_core.c -> > > > > @@ -586,9 +586,6 @@ static int hci_dev_do_close(struct hci_dev *h= -dev) +> > > > > @@ -586,9 +586,6 @@ static int hci_dev_do_close(struct hci_dev *hdev) > > > > > hci_req_cancel(hdev, ENODEV); > > > > > hci_req_lock(hdev); -> > > > > =20 +> > > > > > > > > > - /* Stop timer, it might be running */ > > > > > - del_timer_sync(&hdev->cmd_timer); > > > > > - > > > > > if (!test_and_clear_bit(HCI_UP, &hdev->flags)) { > > > > > hci_req_unlock(hdev); > > > > > return 0; -> > > > >=20 -> > > > > As I said before you need that first hunk to stay for the case wh= -ere -> > > > > there is no device up and you return via the !HCI_UP check. You j= -ust +> > > > > +> > > > > As I said before you need that first hunk to stay for the case where +> > > > > there is no device up and you return via the !HCI_UP check. You just > > > > > moved back to the state before as the stupid timer is active for > > > > > whatever reason even when HCI_UP is not set. -> > > >=20 -> > > > if I read this right then we have the case that we arm this timer f= -or no -> > > > real reason. A device in !HCI_UP should have nothing running. Certa= -inly +> > > > +> > > > if I read this right then we have the case that we arm this timer for no +> > > > real reason. A device in !HCI_UP should have nothing running. Certainly > > > > not the cmd_timer since it will never process any commands. -> > > >=20 -> > > > According to Gustavo, the problem is really in the hci_reset logic = -were +> > > > +> > > > According to Gustavo, the problem is really in the hci_reset logic were > > > > we arm the timer even when shutting down the device. -> > >=20 +> > > > > > The reason why the original patch was sent is, that the timer was > > > running when the thing went out via the !HCI_UP path, which caused the > > > whole thing to explode in the first place. I had no time to figure out > > > why, but moving the del_timer_sync above the > > > if (!test_and_clear_bit(HCI_UP, &hdev->flags)) solved it. -> >=20 +> > > > Oops. Hit send too fast. -> >=20 +> > > > Then it broke the resume on Keith machine and reverting just the hunk -> > which disarms the timer in the=20 -> >=20 +> > which disarms the timer in the +> > > > if (hdev->sent_cmd) { -> >=20 +> > > > path made both scenarios working. So there are two problems: -> >=20 +> > > > 1) Why do we need the del_timer_sync() above the !HCI_UP check ->=20 -> That is still a mysterious to me, the real bug the hiding here. I'm tryin= -g to +> +> That is still a mysterious to me, the real bug the hiding here. I'm trying to > track this down but no luck yet. ->=20 -> >=20 +> +> > > > 2) Why gets the timer rearmed after that ->=20 -> It is armed at each HCI command we send. In the close path we send out an= - HCI +> +> It is armed at each HCI command we send. In the close path we send out an HCI > RESET command that rearms it. -I applied v2 patch from Vin=EDcius that fix all the symptoms. Now we have m= -ore time -to find the real cause of this bug. However I still have no idea, I'm not a= -ble +I applied v2 patch from Vinícius that fix all the symptoms. Now we have more time +to find the real cause of this bug. However I still have no idea, I'm not able to reproduce it. ---=20 +-- Gustavo F. Padovan http://profusion.mobi diff --git a/a/content_digest b/N1/content_digest index 5476bec..4ba8416 100644 --- a/a/content_digest +++ b/N1/content_digest @@ -22,95 +22,81 @@ "* Gustavo F. Padovan <padovan@profusion.mobi> [2011-04-11 19:25:04 -0300]:\n" "\n" "> * Thomas Gleixner <tglx@linutronix.de> [2011-04-12 00:19:32 +0200]:\n" - ">=20\n" + "> \n" "> > On Tue, 12 Apr 2011, Thomas Gleixner wrote:\n" "> > > On Mon, 11 Apr 2011, Marcel Holtmann wrote:\n" - "> > >=20\n" + "> > > \n" "> > > > Hi Thomas,\n" - "> > > >=20\n" - "> > > > > > > > Can the bluetooth folks please have a look at that ASAP? Th=\n" - "e obvious\n" - "> > > > > > > > fast fix for Linus tree is to revert the second hunk for no=\n" - "w, but this\n" + "> > > > \n" + "> > > > > > > > Can the bluetooth folks please have a look at that ASAP? The obvious\n" + "> > > > > > > > fast fix for Linus tree is to revert the second hunk for now, but this\n" "> > > > > > > > needs to be fixed proper.\n" - "> > > > > > >=20\n" - "> > > > > > > Who will submit this patch? I'd rather have your name on it s=\n" - "o that\n" + "> > > > > > > \n" + "> > > > > > > Who will submit this patch? I'd rather have your name on it so that\n" "> > > > > > > people come complain at you...\n" - "> > > > > >=20\n" - "> > > > > > I took a shot at it and just sent a patch (also attached for co=\n" - "nvenience)=20\n" + "> > > > > > \n" + "> > > > > > I took a shot at it and just sent a patch (also attached for convenience) \n" "> > > > > > that should solve the problem.\n" - "> > > > >=20\n" + "> > > > > \n" "> > > > > Aaarg. No. That patch reverts both hunks.\n" - "> > > > >=20\n" + "> > > > > \n" "> > > > > --- a/net/bluetooth/hci_core.c\n" "> > > > > +++ b/net/bluetooth/hci_core.c\n" - "> > > > > @@ -586,9 +586,6 @@ static int hci_dev_do_close(struct hci_dev *h=\n" - "dev)\n" + "> > > > > @@ -586,9 +586,6 @@ static int hci_dev_do_close(struct hci_dev *hdev)\n" "> > > > > \thci_req_cancel(hdev, ENODEV);\n" "> > > > > \thci_req_lock(hdev);\n" - "> > > > > =20\n" + "> > > > > \n" "> > > > > -\t/* Stop timer, it might be running */\n" "> > > > > -\tdel_timer_sync(&hdev->cmd_timer);\n" "> > > > > -\n" "> > > > > \tif (!test_and_clear_bit(HCI_UP, &hdev->flags)) {\n" "> > > > > \t\thci_req_unlock(hdev);\n" "> > > > > \t\treturn 0;\n" - "> > > > >=20\n" - "> > > > > As I said before you need that first hunk to stay for the case wh=\n" - "ere\n" - "> > > > > there is no device up and you return via the !HCI_UP check. You j=\n" - "ust\n" + "> > > > > \n" + "> > > > > As I said before you need that first hunk to stay for the case where\n" + "> > > > > there is no device up and you return via the !HCI_UP check. You just\n" "> > > > > moved back to the state before as the stupid timer is active for\n" "> > > > > whatever reason even when HCI_UP is not set.\n" - "> > > >=20\n" - "> > > > if I read this right then we have the case that we arm this timer f=\n" - "or no\n" - "> > > > real reason. A device in !HCI_UP should have nothing running. Certa=\n" - "inly\n" + "> > > > \n" + "> > > > if I read this right then we have the case that we arm this timer for no\n" + "> > > > real reason. A device in !HCI_UP should have nothing running. Certainly\n" "> > > > not the cmd_timer since it will never process any commands.\n" - "> > > >=20\n" - "> > > > According to Gustavo, the problem is really in the hci_reset logic =\n" - "were\n" + "> > > > \n" + "> > > > According to Gustavo, the problem is really in the hci_reset logic were\n" "> > > > we arm the timer even when shutting down the device.\n" - "> > >=20\n" + "> > > \n" "> > > The reason why the original patch was sent is, that the timer was\n" "> > > running when the thing went out via the !HCI_UP path, which caused the\n" "> > > whole thing to explode in the first place. I had no time to figure out\n" "> > > why, but moving the del_timer_sync above the\n" "> > > if (!test_and_clear_bit(HCI_UP, &hdev->flags)) solved it.\n" - "> >=20\n" + "> > \n" "> > Oops. Hit send too fast.\n" - "> >=20\n" + "> > \n" "> > Then it broke the resume on Keith machine and reverting just the hunk\n" - "> > which disarms the timer in the=20\n" - "> >=20\n" + "> > which disarms the timer in the \n" + "> > \n" "> > if (hdev->sent_cmd) {\n" - "> >=20\n" + "> > \n" "> > path made both scenarios working. So there are two problems:\n" - "> >=20\n" + "> > \n" "> > 1) Why do we need the del_timer_sync() above the !HCI_UP check\n" - ">=20\n" - "> That is still a mysterious to me, the real bug the hiding here. I'm tryin=\n" - "g to\n" + "> \n" + "> That is still a mysterious to me, the real bug the hiding here. I'm trying to\n" "> track this down but no luck yet.\n" - ">=20\n" - "> >=20\n" + "> \n" + "> > \n" "> > 2) Why gets the timer rearmed after that\n" - ">=20\n" - "> It is armed at each HCI command we send. In the close path we send out an=\n" - " HCI\n" + "> \n" + "> It is armed at each HCI command we send. In the close path we send out an HCI\n" "> RESET command that rearms it.\n" "\n" - "I applied v2 patch from Vin=EDcius that fix all the symptoms. Now we have m=\n" - "ore time\n" - "to find the real cause of this bug. However I still have no idea, I'm not a=\n" - "ble\n" + "I applied v2 patch from Vin\303\255cius that fix all the symptoms. Now we have more time\n" + "to find the real cause of this bug. However I still have no idea, I'm not able\n" "to reproduce it.\n" "\n" - "--=20\n" + "-- \n" "Gustavo F. Padovan\n" http://profusion.mobi -d8b5d6fb28e0fb2fcc1205f510679b70030133283d2cce4dd64ef64f35000418 +a150e64a23963770efef31ef79c6830efdca9891a3a96cca9079dd74c0a898c2
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.